THE MICROBIAL ACTIVITIES INVOLVED IN THE ALKALINE FERMENTATION OF SOYBEANS INTO DAWADAWA BY SARAH OPAI-TETTEH A THESIS SUBMITTED TO THE DEPARTMENT OF NUTRITION AND FOOD SCIENCE, UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE AWARD OF MASTER OF PHILOSOPHY DEGREE IN FOOD SCIENCE AI JG US '!' 199 University of Ghana http://ugspace.ug.edu.gh TPZlb W- Wtrc- | University of Ghana http://ugspace.ug.edu.gh DECLARATION I declare that, this w ork was carried out by m yse lf at the Food Research In stitu te (FRI) and the D epartm ent o f Nutrition and Food Science, University o f Ghana, L egon under the Supervision o f Dr. W. K. Amoa-Awua (FRI) and Dr. (M rs.) E sther Sakyi-D awson (Legon) and that it has not been presented in part o r whole to any U niversity fo r the award o f a degree. (Sarah Opai-Tetteh) S tudent (Dr. W. K. Amoa-Awua Supervisor ....... (Dr. (M rs.) E. Sakyi-Dawson) Supervisor University of Ghana http://ugspace.ug.edu.gh DEDICATION Dedicated to my husband Alfred Kofi Darkwa, children M aame Koomah and Paa Kvveku Darkwa and my parents Mr. and Mrs. Opai Tetteh. University of Ghana http://ugspace.ug.edu.gh ABSTRACT This study was initiated to identify the dom inant m icrobial species and investigate their activities during the fermentation o f soybeans to dawadawa. The dom inant m icroorganism s present in both the boiled and roasted soydawadawa, which w ere spontaneously fermented for 72 h, were isolated, characterized and identified using an API kit. The proteinase and a -am y lase activities o f the m icroorganism s w ere determ ined. Quality indices such as pH, moisture, protein and fat contents o f the soydawadawa w ere measured. The breakdown o f proteins during fermentation o f soydawadawa was studied using Gel electrophoresis. The trend in the level o f sugars during fermentation o f soydawadawa was carried out using the Lane and Eynon’s method. A romatic compounds (GC-MS method) produced by the different types o f soydawadawa were also determ ined. Results showed that Bacillus species were the dom inant m icro-organism s found and which increased from 10'’ to 10n cfu/g and 103 to 109 for the boiled and roasted soydawadawa from the start to the end o f fermentation respectively. Lactic acid bacteria ( 104 to 106 cfu/g) were also found. Bacillus subliHs accounted for 48% o f the representative isolates taken from various stages o f fermentation. O ther Bacillus species accounted for 8% to 16% o f the isolates. All the identified Bacillus species demonstrated proteolytic activity in the order Bacillus subtilis >Baciflus furmis > Bacillus cerms > Bacillus pumilus. At a = 0.05, proteolytic activity w as significantly affected by fermentation tim e and production method. However, there was no interaction between these two factors. a -am y lase activity was positively affected by fermentation time, increasing with increase in fermentation time. Hydrolysis o f proteins during fermentation was confirm ed in^soydawadaw a bv an increase in the fraction o f lower molecular weight proteins and more distinct bands. The level o f sugars increased in the first 24 hours but decreased subsequently as ferm entation progressed. A roma compounds produced were more in the roasted products than the boiled. Both products were dominated by 3-Hexanol and 9, 12-Octadecanoic acid. However, Tetradecanoic acid and 1,2-Benzene dicarboxylic acid were found only in the roasted product. University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT 1 am deeply indebted to my supervisors; Dr. (M rs.) E. Sakyi-Dawson and Dr. W. K. A Amoa-Awua for their guidance, helpful suggestions and constructive scrutiny o f this work. 1 am also extremely grateful to M iss Mary Halm, Head o f A nalysis division, FRI, Accra, for her motherly love and help throughout my work. Also to the stafi of the M icrobiology and Chem istry D epartm ents o f FRI, especially Ef'o John, Asiedu, Amoako, Amo, Emma, Kofi, Baisel, Peter, Phillip, A llotey, Amey and Ankrah. I would say God bless you for the assistance rendered to me during my work. I am grateful to Mr. Sackey, Nana and Mr. Kumah for helping me identify my compounds and to all the s ta ff o f FRI, CSIR, Accra, for creating a conducive environment for me to work in. A lso to Mrs. Adiepenah and Sister Francisca o f W omen In A griculture and Development (W IAD) - Accra, I would say thank you. To all my lecturers at the Department o f Nutrition and Food Science, I say thank you for the training given me. This study was facilitated by financial support from DANIDA (The Danish International Development Assistance, Danish Foreign M inistry) and the Government o f Ghana under the collaborative research project “Capability Building for Research into Traditional Fermented Food Processing in G hana” involving the Food Research Institute, Accra, Ghana, the Alfred Jorgensen Laboratory A/S, Copenhagen, Denmark and The Royal Veterinary and Agricultural University, Copenhagen, Denmark I wish to express my sincere gratitude to DANIDA for the assistance. My family cannot be forgotten, especially my husband, children and parents whose regular assistance and encouragement made my study possible. Finally to M iss Joyce Aggrey and all my friends who contributed directly or indirectly towards my study. University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DECLARATION ... ... ... ... i DED ICATION ... ... ... ... ii ABSTRACT ... ... ... ... iii ACKNOW LEDGEM ENT ... ... ... iv L IST OF FIGURES ... ... ... ix L IST OF TABLES ... ... ... x L IST OF APPENDICES ... ... ... xi 1.0 INTRODUCTION ... ... ... 1 1.1 OBJECTIVES ... ... ... 4 2.0 L ITERATURE REVIEW ... ... ... 5 2.1 FERM ENTATION ... ... ... 5 2.2 PARK1A BIGLOGOSA (AFRICAN LOCUST BEAN) ... 6 2.2.1 M icrobiological and Physio-Chem ical changes during the fermentation o f dawadawa ... 7 2.2.2 M icro-organism s involved in fermentation o f daw adawa 2.2.3 Nutritional composition and quality ... 8 2.2.4 Toxicological aspects ... 10 2.3 USES OF SOYBEANS ... ... ... 14 2.3.1 Fermented soybeans ... ... ... 15 2.4 READY TO USE API K IT ... ... ... 18 2.5 ELECTROPHORESIS ... ... ... 19 2.5.1 SDS Polyacrylam ide - gel electrophoresis ... 19 2.6 GAS CH R O M A TO G R A PH Y -M A SS SPECTROM ETRY ... 20 3.0 M ATERIALS AND M ETHODS ... 23 3.1 M ATERIALS ... ... ... 23 3.2 M ETHODS ... ... ... 23 3.2.1 Field Survey ... ... ... 23 3.2.2 Production o f Soydawadawa ... ... 24 PAGE V University of Ghana http://ugspace.ug.edu.gh 3.3 CHEM ICAL ANALYSES A. pH B. Determ ination o f T itrable Acidity C. Proximate analysis i) M oisture ii) Protein content iii) Fat content iv) Ash content v) Carbohydrate content 3.4 M ICROBIOLOGICAL ANALYSES 3 .4 .1 Sampling o f fermenting Soydawadawa 3.4.2 Isolation o f dom inating M icro-organism s 3.4.3 Initial characterisation o f isolates 3.4.4 Maintenance o f isolates 3 4.5 Identification o f Bacillus species 3.4.6 Biochemical Tests for Characterisation o f isolates (i) Protein Hydrolysis (Casein U tilisation) (ii) Starch Hydrolysis (iii)G row th in NaCl (iv)G row th at pH 5.7 (v) Acid production from glucose (vi)Identification using API 50CHB kit 3.4.7 Identification o f Lactic Acid Bacteria i. Initial identification ii. Gas Production iii. G row th at different temperatures iv. Hugh and L eifson’s test v. Identification o f Lactobacillus specics 3.5 DETERM INATION OF SUGARS 24 24 26 26 26 26 27 27 28 28 28 29 29 30 30 31 31 31 31 32 32 32 i -*J J) *» 33 34 34 35 v i University of Ghana http://ugspace.ug.edu.gh 3.6 CHARACTERISATION OF THE PROTEINS OF SOYDAW ADAW A USING GEL ELECTROPHORESIS 3.6.1 Sample Preparation 36 36 3.7 DETERM INATION OF AROMA COM POUNDS ... 37 3.7.1 Sample Preparation and Extraction P rocedures ... 38 3.7.2 Gas Chromatography and Mass Spectrom etry ... 38 3.8 ENZYME ASSAYS 3 .8 .1 Prelim inary Screening o f cultures for Proteinase and a-m y lase activity ... 39 3.8.2 Preparation o f Extraction Buffer for Proteinase activity ... 39 3.8.3 Determ ination o f P roteinase activity ... 39 3.8.4 Preparation o f Extraction Buffer for a -am y lase activity ... 40 3.8.5 Determ ination o f a-am ylase activity ... 40 3.9 DETERM INATION OF AM INO ACIDS IN SOYDAW ADAW A 4 I 3.9.1 Working Solutions ... 41 3.9.2 Procedure ... 44 4.0 RESULTS AND D ISCUSSIONS ... 46 4.1 FIELD SURVEY ... 46 4.2 CHEM ICAL COM POSITION OF SOYDAW ADAW A ... 51 4.3 CHANGES IN pH AND TITRABLE ACIDITY DURING FERM ENTATION ... 52 4.4 M ICROBIAL POPULATION OF FERM ENTING SOYDAW ADAW A ... 54 4.4.1 Aerobic M esophilic count ... 55 4.4.2 Lactic acid bacteria ... 55 4.4.3 Yeasts and moulds ... 59 4.5 PATTERN OF M ICROBIAL GROW TH ... 59 4 .6 OCCURRENCE OF D IFFERENT BACILLUS SPECIES ... 60 v i i University of Ghana http://ugspace.ug.edu.gh 4.7 THE ROLE OF BACILLUS SPECIES IN SOYDAW ADAW A FERM ENTATION ... 63 4.8 CHARACTERISATION OF THE LACTIC ACID BACTERIA POPULATION ... 64 4.9 DETERM INATION OF TOTAL SUGARS ... 66 4.10 ENZYMATIC ACTIVITY IN SOYDAW ADAW A DURING FERM ENTATION 68 4.10.1 Proteinase activity ... 71 4.10.2 a-Amylase activity ... 73 4.11 AM INO ACID PROFILE OF SOYDAW ADAW A ... 77 4.12 CHARACTERISATION OF PROTEIN IN SOYDAW ADAW A ... 81 4.13 AROM A PROFILE OF SOYDAW ADAW A ... 83 5.0 CONCLUSIONS ... 86 6.0 RECOM M ENDATIONS ... 88 REFERENCES ... 89 APPENDICES ... 100 v i i i University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure 2.1 F low diagram o f the processing o f A frican Locust Bean into dawadawa ■ ■ ■ 7 Figure 3.1 Flow diagram for the preparation o f dawadawa from soybean ... 25 Figure 4.1 P icture show ing boiled soydawadawa ... 49 Figure 4.2 Picture show ing roasted soydawadawa 50 Figure 4.3 Total reducing sugars ... 67 Figure 4.4 D iameter o f clear zone on skim milk agar indicating extent o f proteolytic activity o f m icroorganism s ... 69 Figure 4.5 Proteolytic activity in soydawadawa during fermentation ... 70 Figure 4.6 D iameter o f clear zone on starch agar indicating extent o f a-amylase activity o f microorganisms ... 75 Figure 4.7 a-am ylase activity in soydawadawa during fermentation ... 76 Figure 4.8 SDS-PAGE E lectrophoresis show ing protein profile o f soydawadawa at different stages o f fermentation ... 82 P A G E i x University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES P A G E TABLE 2.1 Types o f Dawadawa ... ... 8 TABLE 2.2 Physical and m icrobiological changes during fermentation o f Parkia seeds • • 9 TABLE 2.3 V itam in and toxicant content o f fermented and unfermented dawadawa ... 14 TABLE 4.1 Proxim ate chem ical composition o f soydawadawa ... 52 TABLE 4.2 Changes in pH and titratable acidity (expressed as lactic acid) during the fermentation o f soybeans into dawadawa ... 54 TABLE 4.3 The m icrobiological population o f bacteria and yeast in ferm enting soydawadawa (cfu/g) ... 57 TABLE 4.4 Morphological and biochem ical characteristics o f Bacillus isolates .. 58 TABLE 4.5 Percentage o f identified species from boiled and roasted soydawadawa which fermented various carbohydrates ... 61 TABLE 4.6 D istribution (Percentage) o f different Bacillus species in soydawadawa (final product) .. 62 TABLE 4.7 B iochem ical characteristics o f lactic acid bacteria from boiled and roasted soydawadawa .. 65 TABLE 4.8 Proteolytic activity in soydawadawa during fermentation ... 73 TABLE 4.9 a-A m ylase activity in soydawadawa during fermentation ... 77 TABLE 4.10 Amino acid profile o f fermented and unfermented soybean dawadawa (expressed as mg/aa/g sample) .. 80 TABLE 4.11 M ajor aroma compounds detected during the fermentation o f soydawadawa (GC-MS method) ... 85 x University of Ghana http://ugspace.ug.edu.gh LIST OF APPENDICES APPENDIX 1 APPENDIX 2 APPENDIX 3 Questionnaire Chemical / Reagents used in Amino Acid Analysis Anova on Effect o f fermentation time and type o f soydawadawa on the proteinase activity o f Bacillus species APPENDIX 4 Anova o f Effect o f fermentation time and type o f Soydawadawa on the cx-amylase activ ity o f Bacillus species University of Ghana http://ugspace.ug.edu.gh 1.0 INTRODUCTION In developing countries, the diet o f most people is based on processed cereal grains such as sorghum , rice and maize and on root crops such as cassava and fruits or vegetables such as plantain. These foods by virtue o f the fact that they are consumed in large quantities provide some amount o f protein, yet the quantities are low. Food legumes due to their high protein content generally constitute the natural protein supplement to staple diets (Singh and Rachie, 1985). They are also perfect alternatives for costly meat and fish proteins and help to supplement cereal grains which lack lysine. They are healthy and satisfying, blend well with a full range o f flavours and as the traditional foundation for many ethnic dishes, can give consumers an exotic taste experience (Ihenkoronye and Ngoddy, 1992). Grain legumes are therefore very important in the diets o f the people in most developing countries. Apart from their high protein content, legumes contain good amounts o f B-Vitamins and minerals at adequate levels (Kordylas, 1991). In recent times, there has been a w idespread effort to combat the seemingly persistent problem o f malnutrition among the ever-increasing population particularly in the developing countries. In view o f this, priority has been placed on the need for an effective increase in the exploitation o f the many avenues for a good source o f new and unconventional low -cost and high quality protein foods. Dawadawa is the Hausa name for a strong smelling thick dark coloured paste o f fermented African locust bean seeds (Pctrkia big/obosci). A lthough "Dawadawa" is l University of Ghana http://ugspace.ug.edu.gh used as a food condiment, it contributes to the protein and calorific content of food Steinkraus and Van Veen (1971) stated that traditionally fermented protein rich foods offer excellent possibilities for improving the diets o f people around the world. D awadawa is a good source o f lysine needed to complement cereal foods with inadequate lysine content (Annegers, 1974). It has no cyanogenic glucosides and is nutritionally important (Schery, 1972). Dawadawa is prepared by boiling the seeds o f African locust bean to soften the testa. It is then dehulled, reboiled to soften, drained, fermented and dried. In Ghana, dawadawa has always been processed traditionally from the African locust bean seeds by women in the N orthern Regions o f Ghana. Low- income families use dawadawa as a low-cost meat substitute and they generously add it to soups or stew s and sorghum or millet-based dumplings and porridge. Dawadawa is also popular in several West and Central African countries such as Nigeria where it is also called "iru" and in Burkina Faso where it is known as "soumbala". As the population o f Africa increases there will be more demand for animal protein sources as well as other protein sources such as dawadawa or similar vegetable protein products. The process o f making dawadawa requires little capital but is constrained by the following: 1. The African locust beans whose seeds are used to make dawadawa g row in the wild. 2. The seeds are produced seasonally. 3. Long hours are used in hunting for seeds 4. There is the risk in climbing trees for seeds 2 University of Ghana http://ugspace.ug.edu.gh 5. Process ing o f the seeds into d aw adaw a is time consum ing and labor ious Soybean (Glycine max) like the locust bean is a grain legume, cultivated in many areas o f the world. It has for several centuries contributed significantly to the nutritional requirements o f the people in many East and South East Asian countries form ing a major part o f the traditional diet (Singh el. at. 1987). It is fermented into products such as natto, miso and tempeh and are processed to give soymilk and soy sauce which is a principal flavouring and sauce in Eastern Asia (Carrao el. al. 1994). Soybean is gradually becoming very popular in Ghana as an important source o f protein and o il It contains 40% high quality protein (dry m atter basis) with a good balance o f the essential amino acids closely approximating standards established by FAO (W eigartner, 1987). Its oil, 20-32% (dry matter basis), is quite desirable because it contains a large proportion o f unsaturated fatty acids (FAO, 1989).In recent years the use o f soybeans for producing dawadawa has been introduced into Nigeria and is also being promoted in Ghana by Women in Agriculture and Development (W IAD) and Food Research Institute (FRI). The reasons for promoting soybeans for the production o f dawadawa are that soybeans are: 1. rich in proteins, fat, minerals and vitamins 2. easy to grow 3. obtained all year round. 4. processed into dawadawa which has been found to be acceptable to consumers. Even though a lot o f scientific investigations have been carried out to evaluate microbiological and biochemical changes which occur during the fermentation o f 3 University of Ghana http://ugspace.ug.edu.gh African locust bean seeds into dawadawa very little work has been carried out to explain the fermentation o f soybeans into dawadawa. It is therefore important that the m icroorganisms responsible for the fermentation o f soybeans into dawadawa are evaluated and the extent o f protein modification during soydawadawa production investigated. Such information will help to upgrade the current traditional process used for soydawadawa production and facilitate its promotion and adoption in dawadawa consuming parts o f the country. 1.1 OBJECTIVES This study was initiated to: 1. Identify the dominant microbial species involved in the fermentation o f soybean into dawadawa and characterize some o f their technological properties. 2. Investigate the modification o f protein during soydawadawa production and define the amino-acid profile o f the product. 3. Study the effect o f different processing methods on the composition and quality o f soydawadawa. 4 University of Ghana http://ugspace.ug.edu.gh 2.0 LITERATURE REVIEW 2.1 FERMENTATION Fermentation is a process which relies upon the enzymatic reactions o f micro-organisms normally present or added as starter culture to a food to cause change o f properties, evolution o f heat and effervescence. Although fermentation has long been recognized as one o f the means o f preserving foods, the phenomenon has also been employed to develop desirable characters such as flavour and texture. In the Far East however, fermentation processes used in most food preparations have been found to cause reduction or elimination o f its antinutrients. A traditional food such as tempeh is a well known example (Steinkraus el al. 1960, 1965, 1983; Shallenberger el. al. 1967). Fermentation improves digestibility and increases the concentration o f anti-oxidants and vitamins. The micro-organisms involved in fermentation are not only catabolic, breaking down several complex vitamins and other growth factors (Potter, 1968). The industrial production o f such materials as riboflavin, vitamin B2 and the precursor o f vitamin C are largely by special fermentation processes. Fermentation also causes the liberation o f nutrients locked into plant structures and cells by indigestible materials. Certain bacteria, yeasts and moulds break down indigestible protective coatings and cell walls both chemically and physically. These micro-organisms during fermentation release certain enzymes which split cellulose, hemicellulose and related polymers which are not digestible by man into simpler sugars and sugar derivatives (Potter, 1968). Fermentation could either be in the liquid state or solid slate. Solid substrate fermentation is defined as the growth o f micro-organisms on solid materials without the presence o f free liquid. This covers the fermentation o f plant derived solids to produce foods for human consumption. In the case o f the liquid substrate fermentation, there is the presence o f free liquid (Paredes-Lopez and Harry, 1988). 5 University of Ghana http://ugspace.ug.edu.gh 2.2. PARK I A BIGLOBOSA (AFRICAN LOCUST BEAN) The taxonomy o f African locust bean trees has been in a state o f flux until recently. Because o f this, various reports in the literature have referred to this same tree as Parkin fi/icoidea, Parkia. bicolor and Parkict. biglohosa. The related species Parkin. fiUvoidm is indigenous to the forests o f East and Central Africa. The Parkia. bicolor is found in the forest regions o f West Africa. The fruit provides a constant source o f valuable protein in the dry season. The Locust bean tree is also used for medicinal purposes and as a source ol mouth wash to relieve toothaches. The bean husks (seed coats) are used with indigo dye to improve the lustre o f fabrics, while the tree bark yields a red tannin for dying leather. The dry yellow powdering pulp is rich in sweet carbohydrate and can be mixed with cereal, meat or soup. It can also be made into candy or be pressed into cakes for preservation (Abbiw, 1990). Dawadawa is the name given to fermented African Locust bean seeds (Parkia species) and is an important condiment in the Northern Regions o f West Africa. Dawadawa as known often by the Hausa name has other names such as ini in Yorubaland o f Nigeria, Ogiri-iga/a in Iboland, Kpalugu among the Kusasis and Dagombas o f northern Ghana, Kotigo and Tsogo in Ghana; Kinda in Sierra Leone and neleion or soumbala in Gambia and other French speaking countries. (FAO, 1989; Campbell-Platt, 1980; Ihekoronye and Ngoddy, 1985). The steps involved in preparation o f dawadawa from African locust bean seeds are shown in Figure 2.1. The process is believed to involve a number o f proteolytic changes in the substrate due to bacterial action. The fermentation which is by chance inoculation is by various sub-species o f the Bacillus subtilis group (Odunf'a, 1981, FAO, 1989). The African locust bean seeds are boiled in water to soften the seed coat or testa and then dehulled by pounding in a mortar or by rubbing them between the palms. The cotyledons are washed and boiled further for 1-2 h. These are spread in trays or baskets and wrapped in many layers o f ju te sacks or leaves. The seeds ferment for about 36 h or longer. The fermented seeds have a greyish sticky mucilage covering and a strong ammoniacal smell. The colour also changes from light brown to dark brown. 6 University of Ghana http://ugspace.ug.edu.gh African locust bean seeds 1 Boil in water(18 - 24 hours) 1 Cool 1 Dehull by pressing bewteen palms etc I Wash - Discard seed coats and undehulled beans 1 Boil cotyledons in water (1 - 2 hours) 1 Drain and pack in baskets, perforated pots or spread on trays 1 Cover cotyledon with leaves or ju te bags 1 Ferment for 3 - 4 days I Air and /or sun dry 1 Dawadawa Fig. 2.1 Flow diagram of the processing of African locust bean into dawadawa. University of Ghana http://ugspace.ug.edu.gh Different ethnic groups process dawadawa from different raw material as shown in Table 2-1 Table 2-1 TYPES OF DAWADAWA Ethnic group Raw material Shape Bean Dagarti Locust Bean Ball Paste Soybean Dagomba Locust Bean m ixture or Cylindrical Ball and Flat Bean Locust Bean Paste Frafra Locust Bean Ball Bean Paste Kussasi Locust Bean Soybean Cylindrical Ball and Flat Bean mixture or Locust bean Paste groundnut mixture Mossi Locust Bean Ball Bean Source: Campbell - Platt (1980) 2 .2 .1 Microbiological And Physico-Chemical Changes During The Fermentation of Dawadawa Dawadawa fermentation is a solid-substrate fermentation where the growth o f micro­ organisms occur without the presence o f free liquid. The physiological and microbiological changes that occur during fermentation o f African locust bean seeds into dawadawa arc presented in Table 2-2. University of Ghana http://ugspace.ug.edu.gh Table 2-2 Physical And Microbiological Changes During Fermentation O f Parkin Seeds Fermentation time (Ii) Moisture Content (% ) Temp. (°C) pH Plate Aerobic Count Anaerobic 0 43.0 25 7.0 0 5.0x10’ 12 48.0 30 7.2 0 2.4x1 ( f 24 54.0 42 7.5 1.8.\10'1 1.2x10'’ 36 56.0 45 8.1 2.5x10s 3.0x104 (Odunfa 1983) Several physico-chemical changes occur during fermentation o f African locust bean seeds The temperature o f the fermenting seeds increase from about 30 "C to a maximum o f about 50 °C in 24h. The pH increases to about 8.1 during the first 30h o f fermentation, due in pan to the production o f ammonia. The most significant biochemical change that occurs during dawadawa fermentation is protein hydrolysis. This is due to the high proteinase activity which results in rapid amino acid production (Odunfa, 1983). Odunfa (1983) studied the extracellular enzymatic activities o f the micro-organisms in fermenting African locust beans. The enzymes a -galactosidase, |3- galactosidase, sucrase, proteinase, amylase and lipase were detected in fermenting dawadawa. Glycosidases hydrolyze African locust bean oligosaccharides (stachyose, rafiinose and sucrose) and other complex sugars during fermentation. The high temperature which develops during the fermentation hastens enzymatic activities and biconversion. The optimum activity o f the glycosidases occur at 12h and 24h o f fermentation respectively. Invertase activity is highest at 36 h. whilst that o f amylase is fairly low and evident only in the first 24h o f fermentation 9 University of Ghana http://ugspace.ug.edu.gh (Odunfa, 1983). Presumably all the starch is hydrolyzed during this period and is not detected in fermented dawadawa (Watson, 1971). The sugars produced during fermentation through amylolysis provide easily utilizable substrates for the micro-organisms. Although oil constitutes 31 to 40% o f the locust bean seeds, lipase activity is low in fermenting African locust beans and there is fluctuation in activity throughout the fermentation. Campbell-PIatt (1980), suggested that lipolytic activity occurred in the later stages o f African locust bean fermentation. This observation was based on the increase in the number o f lipolytic micro-organisms. Another significant change during the fermentation o f African locust bean seeds is a decrease in percentage o f free fatty acids i.e. from 0.6% in the cooked unfermented African locust bean seeds to 0.1% in dawadawa (Odunfa, and Adesomoju, 1983). The decrease in free fatty acids is desirable since large amounts o f free fatty acids in foods can result in objectionable taste and cause rancidity. However they sometimes produce characteristic flavours in some foods. 2.2.2 Micro-Organisms associated with Dawadawa fermentation Only bacteria have been reported to be associated with the dawadawa fermentation Ikenebomeh (1982), reported a few fungi in the dawadawa fermentation as contaminants Campbell-PIatt (1980), reported that PeniciIlium, fermenting oval budding yeasts, and the film yeast Candida were incidental contaminants. They constituted 3% o f the total microbial isolates. The majority o f the bacteria found in dawadawa are aerobic while approximately 10% are anaerobic after 36h.of fermentation (Table 2-2), 10 University of Ghana http://ugspace.ug.edu.gh Odunfa (1981), first reported that the predominant fermentation micro-organism was a Bacillus, possibly Bacillus subtilis and other species. Later Odunfa and Adesomoju (1983), confirmed the presence o f Bacillus pumilis, B. Hcheniformis and B. suhtilis in the fermentation. Similar findings have been reported elsewhere (Campbell-Platt 1980; Ikenebomeh, 1982). In some Nigerian samples, Peiliocuccus and two strains of Staphylococcus saprophyiicus were detected (Odunfa, 1981). Adewuyi (1983), confirmed the presence o f Bacillus subtilis strains in dawadawa fermentation. He reported that Bacillus subtilis is responsible for production o f acceptable dawadawa. All isolates o f the Bacillus subtilis from dawadawa were found to be proteolytic, with only a few strains being amylotytic (Adewuyi, 1983). Campbell-Platt (1980), found that Bacillus species composed 95% o f both proteolytic and amylolytic isolates, and 76% o f lipolytic isolates. They were thermotolerant, grew at 50°C with an optimum temperature at 35"C to 40°C. They were facultative anaerobes and grew over a wide range o f pH. Diawara et. al. (1998) found the dominating Bacillus species in dawadawa produced in different parts o f Burkina Faso to be Bacillus subtilis. 2.2.3 Nutritional Composition And Quality Campbell-Platt (1980) reported that on a moisture free basis, dawadawa contains 38.5% protein, 3 1.2% fat, and 23,6% carbohydrate, compared to unfermented African locust bean seeds, which have 30% protein, 15% fat and 49% carbohydrate. Fetuga et. al. (1973) observed a small decrease in sulfur-containing amino acids and a greater decrease in aspartic and glutamic acids as a result o f fermentation o f locust beans. Like many dry beans, locust beans are low in the sulfur-containing amino acids, cysteine and methionine (Fetuga et. al., 1973). Eka (1980) found dawadawa to be low in the essential amino acids. 11 University of Ghana http://ugspace.ug.edu.gh leucine, isoleucine, phenylalanine, and tryptophan. However, the essential amino acids in the main meal in which dawadawa is used as a condiment help to complement the low levels in dawadawa. Unsaturated fatty acids account for about 60 to 80% o f total lipids in dawadawa (Busson, 1965; Girgit and Turner 1972). Linoleic acid is the major fatty acid in dawadawa and others found in appreciable amounts include palmitic, stearic and oleic acids (Odunfa, 1983; Girgit and Turner, 1972). Much o f the available reducing sugars and other carbohydrates are utilized by micro-organisms during fermentation. The raffinose family o f oligosaccharides and sucrose decreased significantly during fermentation (Odunfa, 1983). Fully fermented dawadawa contains very little reducing sugars. Watson, (1971) found less than 0.1% glucose, 0.3% fructose and no starch or sucrose in fermented dawadawa. Dawadawa contains appreciable amounts o f folate (Keshinro, 1983; Hug el. cil. 1983). Out o f 24 foodstuffs analysed, dawadawa was the second highest in folate content after sweet potatoes (Hug el, al. 1983). Folate content in dawadawa ranges from 0.89 to 0.95 (.ig/g on a dry weight basis. Significant amounts o f folate can be derived from dawadawa to satisfy the adult RDA requirement for folate. Dawadawa contains most o f the important minerals adequate to meet the RDA requirements with the exception o f calcium which is deficient in the diet o f many West Africans (Oke, 1972). The digestibility o f fermented beans averages 97.6% with a range o f 93.2 to 99.4% (Umoh and Oke, 1974). The digestibility o f protein measured as protein efficiency ratio (PER), net protein utilisation (NPU), and biological value (BV) o f 12 University of Ghana http://ugspace.ug.edu.gh fermented African locust beans were found to be higher than that o f the raw seeds (I'etuga et. al. 1973; Um ohandO ke, 1974). 2.2.4 Toxicological Aspects The levels o f toxic substances such as oxalic acid, phytic acid and hydrocyanic acid are high in unfemiented locust beans. However, some o f these toxic substances are reduccd during cooking and fermentation o f dawadawa (Table 2-3). Although hydrocyanic acid is present in African locust beans, its level is not harmful to the body (Oke, 1969). Soluble oxalate is present in low levels in dawadawa and is toxic to animals. There is an observed decrease in the phytic acid and oxalic acid content o f dawadawa during fermentation which makes more mineral elements available (Eka, 1980). Phytohaemagglutinins in various species o f Parkia are normally destroyed by cooking and are not found in fermented dawadawa. Studies by Alozie et al. (1980) showed that aflatoxins are not present in dawadawa and is attributable to the alkaline pH which is unfavourable to mould growth. 13 University of Ghana http://ugspace.ug.edu.gh Table 2-3 Vitamin and Toxicant Content of Fermented and Unfermented Dawadawa (Eka, 1980; Campbell Platt, 1962) Component Fermented Dawadawa Unfermented Dawadawa Thiamin (mg/!00g) 1.35 065 Riboflavin (mg/lOOg) 1.30 0.45 Niacin (mg/lOOg) 5.30 7.50 Oxalate (g/lOOg) 0.12 0.21 Phytic acid P (g/1 lOg) 7.50 15.00 2.3 USES OF SOYBEAN While the use o f soybeans for food has a long history in china, the far East and Southeast Asia, it is still at an incipient stage in other regions o f the world. In this connection, it is a pleasing tendency that with peoples increasing awareness o f health and their changing dietary habits, they are becoming more interested in soybean products. Considering the value o f soybeans as a protein source and ease o f handling as a material, a campaign for more soybean consumption as human food will play a very important role in solving the world's food issues (Horii, 1997). Horii (1997), found that soybeans had the effect o f controlling the accumulation o f fat in the liver and the body. He also found that linoleic acid which accounts for about 50% o f soybeans fatty acids together with linolenic acid have the function o f reducing the amount o f cholesterol that settles on the wall o f vessels. In addition, the phospholipids, mainly lecithin which is high in soybeans, prevent the 14 University of Ghana http://ugspace.ug.edu.gh settlement o f cholesterol on vessel walls. Because soybeans contain these substances, they are expected to help control the occurrence o f hypertension. 2.3.1 Fermented Soybeans Lee et. at. (1983) investigated the natural fermentation o f soybean for 7days at ambient temperature. They found out that, the content o f riboflavin increased from 98 to 309.4 (.ig/100g dry matter, relative nutritive value from 78.66 to 94.59% and available lysine from 6.56 to 7.38 mg/gN respectively. Also during fermentation, the activities o f protease and lipase increased with progress o f proteolysis during fermentation. Fermentation o f soybean into a dawadawa type o f product using whole cotyledons in four days was undertaken by Barimalaa el al. (1994). Isolates from fermenting cotyledons showed Bacillus subtilis and Bacillus Hchenformis as the major fermenting micro­ organisms. The fermentation increased moisture, protein and fat contents o f cotyledons. Total available carbohydrate reduced in 48h to less than 50% o f value at the start o f fermentation but then increased to 92% on the fourth day. The increase was attributed to secretion o f slime by fermenting micro-oganisms. Fermentation also reduced the trypsin inhibitor activity o f cotyledons. Several articles in the literature make reference to the clinical significance o f Bacillus species in the food industry. However Bacillus species mainly Bacillus subtilis have been identified as the main organisms responsible for the alkaline fermentation o f legumes such as the African locust bean seeds, melon seeds, oil bean, sesame seeds, castor oil bean and flutted oil bean into tradtional products in West Africa (Odunfa and Oyewole 1986; Antai and Ibrahim 1986; S teinkraus 1991). The fermentation o f soybeans into natto, thua-nao and Kinema in Asia are also reported to be carried out by Bacillus subtilis (Suzukin and Ohta 1979; Campbell Platt, 1987; S teinkraus 1991). 15 University of Ghana http://ugspace.ug.edu.gh Nikkuni (1997) reported a conspicuous and charming characteristic o f fermented foods from soybean. He reported that tempeh, miso and other fermented soybean foods have some antioxidant substances which are non-existent in raw soybeans. Among the important types o f fermented foods are tempeh, soy sauce or “shoyu”, miso, “sufu” or Chinese ciieese and natto. Tempeh, a popular food in Indonesia and Malaysia, is a fermented soybean product made by the action o f Rhizopus ohgosporus on cooked and dehulled soybeans. The dehulled and boiled soybeans are innoculated with some tempeh from a previous fermentation, and finally wrapped in banana leaves and allowed to ferment until the mycelium o f the tempeh mould has grown over and through the soybeans to make a solid mass. Soy Sauce represents one o f the largest uses o f soybeans in the Orient and is used extensively as a condiment. It is a dark brown liquid, with a pleasant aroma used primarily as a flavoring agent. Its high salt content o f about 18% makes it an adjunct for many bland foods with which it is used. "Tamari" is a soy sauce made in China, in which the proportion o f soybeans is higher than regular soy sauce. "Chau yan" is a Chinese name and “toyo” the Phillipine name for this condiment. In Indonesia soy sauce called “Ketjap” is made from black soybeans. Fermentation is dependent on three micro-organisms a mould Aspergillus oiyia , a yeast Zygosaccharoniyces soja and a bacterium Ixiclobacillus delbrueku Miso is a fermented food product prepared in Japan, China, Taiwan, Phillipines, Indonesia and other countries in the Orient. It is essentially a fermented blend o f rice, soybeans and sometimes barley or malt. It has the consistency and colour o f peanut butter. It is used as a flavoring substance for foods and as a spread for bread. A two-stage fermentation is used, the first, an aerobic fermentation by strains o f Aspergillus oryzae and the second, an anaerobic fermentation carried out by Saccharomyces rouxii. A mixture o f strains o f Aspergillus oryzae are innoculated into steamed rice and fermented for 48 h at 4 0 T Before spore fermentation occurs, mycelial growth is arrested and this is known as the Koji Simultaneously soybeans are washed, soaked, steamed and cooked. These are blended with salt in the proportion o f 4 parts moulded rice or Koji, 10.4 parts soybean, 2 parts salt and 1 part old miso and water. This blend is fermented aerobically for 7 days at 28 ”C and then two months longer at 35 °C. It is then placed in a vat where the second fermentation, 16 University of Ghana http://ugspace.ug.edu.gh essentially anaerobic, continues for two months after which it is allowed to age lor two weeks at room temperature (Pederson, 1971). Several types o f miso are made "Kome miso" made with soybeans and milled rice, “Mugi miso” made with soybeans and barley, and “Maine miso” made with soybeans alone, Sufu is prepared by mould fermentation o f cakes o f finely ground precipitate soybeans. Soybean milk is cooked to eliminate the bean-like flavour and then pressed, cut into cubes called “tofu”, sprayed with acid-saline solution, innoculated with a mould culture, usually Mucor spp. and incubated at 12-20 °C for 3-7 days. The mouldy cubes known as phetzes are placed in a solution o f 12% NaCI and 10% ethanol and then aged up to 2 months. "Fuyo", "Fu-ju" and "Tsofu" are other names applied to Sufu. Natto is known as a highly nutritious food containing protein, fat and various vitamins (Ueda, 1989). During natto fermentation, soybean protein which has been denatured by cooking process, is hydrolyzed by proteases produced by Bacillus subtilis into peptides and amino acids (Ueda, 1989). The percent liberation o f amino acids i.e. the ratio o f free amino acid content to the total amino acid content is 8 0% for natto and I 1% for soybeans (Nikkuni et. al. 1995). Natto is known as a highly nutritious food containing protein, fat and various vitamins (Ueda, 1989). Recently, natto is also attracting much attention as a food having several physiological functions including the action o f lowering blood pressure (Sumi, 1990). 17 University of Ghana http://ugspace.ug.edu.gh 2.4 READY - TO -USE API KIT In the present work, comprehensive carbohydrate utilization profiles o f isolates were determined in a ready-to-use API kits in addition to classical biochemical tests used to characterize and identify microbial isolates. The biochemical profile o f Bacillus species were determined in API 50 CH kit (BioMerieux SA) using API 50 CHB media (BioMerieux SA). This allowed the utilization o f 49 carbohydrates by each tested strain to be studied. In the API kit, the carbohydrates substrates are present in a dehydrated form and are rehydrated with the inoculated ready to use medium. Fermented carbohydrates produced a decrease in pH which was detected by a colour change o f the incorporated indicator. The result constituted a biochemical profile o f the strain and the species could be identified by comparison to a reference table or a database. API tests have been found to give more reproducible results than the classical tests in the identification o f Bacillus species. (Logan and Berkerley 1981). The carbohydrate substrates present in API 50 CH include glycerol, erythritol, D-arabinose, L-arabinose, ribose, D-xylose, L-xylose, adonitol, f3 methyl-xyloside, galactose, D-glucose,D-fructose, D-mannose,L-sorbose, rhamnose, dulcitol, inositol, mannitol, sorbitol, a methyl -D - mannoside, a methyl -D-glucoside, N acetyl glucosamine, amygdaline and arbutin. The rest are esculin, salicin, cellobiose, maltose, lactose, melibiose, saccharose, trehalose, inuine,melezitose, D-raffinose, amidon, glycogen, xylitol, P gentiobiose, D-turanose, D- lyxose, d-tagatose, D-fucose, L-arabitol, glucanate, 2 ceto-gucanate and 5 ceto-gucanate 18 University of Ghana http://ugspace.ug.edu.gh 2.5 ELECTROPHORESIS Electrophoresis is concerned with the migration o f charged components within a fluid medium under the influence o f an electric field. The charged components may be molecular, for example proteins, or cells or even charged particles. Similarly, the fluid medium may be liquid or gaseous, but in biological systems it is only electrophoresis in the former that is o f interest (Gordon and Macrae, 1992). The basis o f electrophoretic separation is that molecules with different charges or sizes will migrate at different velocities. If there is an interaction between the charged molecules and the support material, then the situation may be complicated further. Support material is used to reduce other effects such as convection, which would disrupt the migration o f the components as discrete bands (Gordon and Macrae, 1992). Electrophoresis o f the total cellular proteins in polyacylamide gels provide a partial separation in which individual bands mostly represent several proteins. 2 5 1 SDS Polyacrylamide-Gel Electrophoresis. The original electrophoretic systems used to determine protein fingerprints involved polycrylamide rod gels used in non-denaturing conditions, but more recently sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) has found greater application (Priest and Austin, 1993). SDS-PAGE is a low cost reproducible and rapid method for quantifying, comparing and characterizing proteins. It separates proteins based primarily on their molecular weight. SDS binds along the length o f the polypeptides chain and the length o f the reduced SDS-protein complex is proportional to its molecular weight. The mobility o f charged molecules in gels is a function o f both their size and charge, so it is 19 University of Ghana http://ugspace.ug.edu.gh quite possible for molecules with the same electrophoretic mobility (under specified conditions) to be quite different in terms o f relative molecular mass. Proteins may associate in buffer solutions and hence migrate during electrophoresis as aggregates thereby not reflecting their true charge or relative molecular mass. To overcome these problems, a method which includes an anionic detergent, such as sodium dodecyl sulphate (SDS), into the buffer system is used. SDS binds to hydrophobic sites within the protein, hence reducing the possibility o f hydrophobic bonding between protein molecules. It also imparts a large negative charge to the protein units, its contribution largely swamping any ell'ect o f charged groups within the protein. Under these conditions, it has been found that the electrophoretic mobility o f a protein is inversely related to the log o f its relative molecular mass (Gordon and Macrae, 1992). 2 6 GAS CH RO M A TO TOG RA PH Y -M A SS SPEC TR O M ETR Y (G .C -M S) Food products comprise trace levels o f numerous flavour components dispersed in a highly complex and often nonhomogenous matrix. To enhance the sensitivity o f analytical methods preconcentration is important. A number o f techniques are commonly employed for extracting and concentrating flavour components from products (Jennings, 1980: Bemelmans, 1981: Teranishi and Kint, 1993). They include solvent extraction, headspace sampling and distillation. Each technique has its own advantages and disadvantages and often compliment one another. To choose an optimal method one must consider the analytical objectives, type o f products and the type o f flavour constituents likely to be present. The first o f these tasks is generally carried out using GC coupled with mass spectrometry. 20 University of Ghana http://ugspace.ug.edu.gh Gas Chromatography is ideally suited to the analysis of volatile components in the environmental o r in the headspace above biological materials, Analysis of volatiles is important in various areas including determination o f flavour and odour components of foods (Williams, 1971). Gas Chromatography achieves separation o f mixtures by partition o f components between a mobile gas phase and a stationary phase. The stationary phase could be an involalile liquid coated onto an inert solid support (Gas Liquid Chromatograpy) or comprises particles o f a solid absorbent (Gas Solid Chromatography). The time taken for a molecule to pass through a GLC column is known as the retention time, tr, and is dependent on the partition coefficient k.. K= mass o f vapour dissolved in unit column length o f stationary phase mass o f vapour dissolved in unit column length o f mobile phase. The separation o f two components in a mixture is dependent on the difference in their retention times (Dtr) and the mean peak width (Wb) o f the bands. GLC has become a major analytical technique which gives excellent separation o f components from many complex mixtures (Gordon and Macrae, 1992). Mass spectrometry is a powerful technique for the identification o f pure compounds. It can also be used for confirmation o f the purity o f a sample and for quantitative analysis of' mixtures. The mass spectrum consist o f a series o f peaks o f varying intensity plotted against the mass-to-charge ratio (m/y),Mass spectrometry combines high specificity with great sensitivity, since amount o f less than a picogram o f some compounds can be detected 21 University of Ghana http://ugspace.ug.edu.gh GC-MS represents a very powerful combination o f separation and structural identification techniques. It is required for many non-routine studies involving the analysis o f complex mixtures. The advantages o f a mass spectrometer are that it provides a sensitive specific and universal method o f detection. Full structural identification o f unknown components is often possible from the mass spectrum. GC-MS rely on the separation o f mixtures into individual components which can readily be identified by mass spectrometry. Packed column GC-MS is best performed by enrichment o f the eluent with a molecular separator. This device selectively removes carrier-gas molecuies from the gas flow entering the mass spectrometer (Gordon and Macrae, 1992). 22 University of Ghana http://ugspace.ug.edu.gh 3.0 MATERIALS AND METHODS 3.1 M A TER IA LS Soybean ((Jlyciim max) was purchased from Makola Market, Accra-Ghana. M icrobiological media Plate count agar (PCA) Difco 0479-17-3, Detroit U.S. A Man, de Rogosa and Sharpe agar (MRS) Merck 1,10660 Darmstadt, Germany. Malt agar (MA) OxoidCM59, England Yeast extract Difco, 0127 1 7 -9 , U .S.A Bacteriological peptone - Oxoid 1375, England. Agar - Sigma 91H01625, U .S .A Nutrient agar - Difco Detroit, U .S.A Nutrient broth - Difco 36201JA 3.2 M ETH O D S 3.2.1 Field Survey A brief field study was carried out in Nima, M ateheko and Madina all suburbs o f Accra to select two major sources o f traditional dawadawa production. The field study involved the use o f questionnaires and informal interviews o f dawadawa sellers. Information gathered included the following: types o f dawadawa available, raw materials for its preparation and its availability, methods o f production, ingredients used, packaging, marketing and consum ers’ perception o f good quality dawadawa (Appendix I). 23 University of Ghana http://ugspace.ug.edu.gh Two experienced dawadawa producers were selected to produce the two different types o f dawadawa used as control for laboratory studies. All other experimental work were carried out at the Food Research Institute, Accra, Ghana and at the Department o f Nutrition and Food Science, University o f Ghana, Legon. 3.2.2 Production of Soydawadawa Two different types o f soydawadawa were prepared as shown in the flow diagram (Figure 3-1). One type was made from boiled soybeans and the o ther from roasted soybeans as shown in Figure 3.1 Soybean seeds were cleaned, boiled for 30 min cooled and dehulled by pressing between palms or by pounding in a mortar. The dehulled beans w ere washed to get rid o f the seed coat and undehulled beans. It was further boiled for lh, drained and packed into baskets lined w ith banana leaves. This was covered with more leaves and left to ferment for 72h to obtain boiled soydawadawa. For the roasted soydawadawa, the soybean bean seeds, after cleaning were roasted for 30 min, cooled and dehulled and further treated as in the boiled one. 3.3 CHEM ICAL ANALYSES A) pH Fermenting daw adawa samples weighing lOg were homogenized in a blender with 90ml o f distilled w ater and the pH determ ined with a pH meter (PHM series Lab. pH meter; PHM 92, Radiometer, Copenhagen, Denmark). 24 University of Ghana http://ugspace.ug.edu.gh Figure 3. Soybean 1 Clean Roast for 3D min 1 Boil for 30 mins f Cool Dehull 1 Wash 1 Boil cotyledons in w ater ( lh ) I Drain through sieve .1 Spread in baskets lined with leaves 1 Ferment for 2-3 days 1 Mould into various shapes 1 Air and/or dry 1 Dawadawa I: Flow diagram for preparation of dawadawa from soybean 25 University of Ghana http://ugspace.ug.edu.gh B) T ITRATABLE ACIDITY OF DAWADAWA Titratable acidity was determ ined by the titration o f 80ml o f filtrate obtained from lOg o f dawdawa samples macerated in a blender in 250ml distilled water, against 0 . 1N NaOH with 1% phenolphthalein. 1ml o f 0.1 N NaOH was taken as equivalent to 9.008 x 10'"'g lactic acid. C) PROXIMATE ANALYSES i) Moisture M oisture content was calculated by drying the well-mixed sample at 130.°C to constant weight (AOAC 1990). For all samples, 2g were weighed into moisture dishes and dried in an air oven set at 130 °C to constant weight. Dishes w ere cooled in a desiccator and reweighed. The percentage moisture was calculated as: % moisture = Loss in weight x 100 initial weight o f sample This was calculated on D ry M atter Basis ii) Protein Content The protein content o f samples was determ ined by multiplying total nitrogen, estimated by micro-kjeldahl method, by 5.7 (AOAC 1990). For all samples, 2g were weighed on filter paper and digested with concentrated sulphuric acid in a kjeldahl flask. It was then distilled into 2% boric acid and back titrated with 0 , 1M HCL with methyl red as indicator. Protein content was calculated as: % Protein = % N itrogen x 6.25 This was calculated on Dry M atter Basis. 26 University of Ghana http://ugspace.ug.edu.gh iii) Fat Content Fat content was determ ined by ether extraction using the Soxhlet-petroleum ether extract method (AOAC 1990). For all samples, 2g were weighed into a soxhlet paper thimble, covered with cotton wool and placed in the extraction unit o f the soxhlet apparatus. A weighed clean oven dried soxhlet flask was attached to the lower end o f the extraction unit which was filled with petroleum ether and heated for some hours. The percentage o f fat was calculated as: % Ether extract = Increase in Flask weight x 100 weight o f sample This was calculated on D ry M atter Basis iv) Ash Asli content was measured by heating the sample at 600°C until the difference between the successive weighings was (less than or equal) < ling (t\bAC 1990). For all samples about 2g o f the dried sample was transferred into a weighed crucible (Silica dishes) and heated for about 2 h to red hot (600°C) in a furnace. The fine ash was weighed after cooling in a desicattor to give the ash content. % Ash = Weight o f Ash x 100 Dry sample weight This was calculated on D ry M atter Basis 27 University of Ghana http://ugspace.ug.edu.gh v) Carbohydrate Carbohydrate content was calculated by difference: % Carbohydrate = 100 - (% M oisture + % Ash + % Fat + % Protein This was calculated on D ry M atter Basis 3.4 M ICROBIOLOGICAL ANALYSES 3 .4.1 Sampling o f Fermenting Soydawadawa The tw o types o f dawadawa were sampled in duplicate on 4 separate occasions precisely at Oh, 24h, 48h and 72h o f fermentation. Samples o f lOg o f soydawadawa at 0, 24h 48h and 72h o f fermentation w ere aseptically collected into stom acher bags (Seward Medical, London, England) for analysis. Soydawadawa samples were taken from within the fermenting mass after the surface layers had been removed aseptically using a sterile spatula. For all samples, lOg were added to 90ml sterile diluent containing 0.1% peptone, 0 .85% NaCl, with pH adjusted to 7.0 and homogenized in a stom acher (Lab Blender, Model 4001, Seward Medical) for 120s at high speed. From appropriate ten-fold dilutions, enumeration o f aerobic mesophiles was carried out on Plate Count Agar (PCA, Difco) incubated at 37°C for 3 days. Lactic acid bacteria were enumerated on de Man, Rogosa and Sharpe Agar (MRS Merck) incubated anaerobically in an anaerobic ja r with anaerocult A (M erck) at 37°C for 4 days. Mould and Yeast counts were enumerated on Malt Agar (MA, Oxoid) containing lOOmg chloramphenicol 28 University of Ghana http://ugspace.ug.edu.gh (Chloramphenicol selective supplement Oxoid) and 50mg chlotetracycline (sigma C- 4881, St. Louis, USA) per litre and incubated at 30°C for 7 days. 3 4 2 ISO LA T IO N O F D O M IN A T IN G M IC R O O R G A N ISM S The highest dilution plate or suitable plate with a total o f about 30 colonies in each quadrant was used for the isolation. Distinct individual colonies were picked with a sterile innoculating loop and subcultured in the corresponding broth medium and streaked on to the agar substrate repeatedly until pure cultures were obtained. 3.4 3 IN IT IA L C H A R A C TER IZA T IO N O F ISO LA TES Bacillus snecies Isolates from PCA w ere subcultured in nutrient b ro th/agar and Bacillus species were characterised using morphological examination and biochemical test comprising o f colony and cell morphology, Gram reaction and catalase production. The proportion o f Bacillus species in the isolates was used to calculate the total numbers o f the species present in the sample. Gram reaction was carried out according to a modification o f the method o f Lillie (1928) by Parry el al. (1983). For catalase production a loopful o f culture was mixed into a drop o f 3% hydrogen peroxide on m icroscope slide and observed for the production o f gas bubbles to indicate production o f catalase. Oxidase test was carried out using strips o f Oxidase paper. (Lillie (1928) by Parry et. al. 1983) 29 University of Ghana http://ugspace.ug.edu.gh Lactic acid bacteria Isolates from MRS w ere subcultured in MRS medium and examined by Gram reaction, catalase production, oxidase test, aerobic and anaerobic grow th, colony and cell morphology. Y easts and m oulds Isolates from MA were subcultured in yeast-m alt-peptone-glucose broth and agar (M YPG) containing 3g, yeast extract (Difco) bacteriological peptone (Oxoid) lOg, g lucose (M erck) with or w ithout 15g, agar (sigma) and examined by colony and cell morphology. 3 .4 .4 MAINTENANCE OF ISOLATES Plate Count Agar (PCA) isolates were maintained on PCA slants stored at 4°C. Colonies isolated from de Man, Rogosa and Sharpe (MRS) Agar w ere preserved by subculturing in MRS broth and vortexing I ml o f culture broth with 1ml o f 50% glycerol in a C ryotube and frozen immediately at 40°C. Malt Agar (MA) isolates were maintained on M YPG agar slants stored at 4°C. 3.4.5 IDENTIFICATION OF BACILLUS SPECIES Bacillus species were recognised from initial tests o f PCA isolates as Gram positive, catalase positive rods with phase bright spores. The species o f BaaIIu\ isolates were identified according to Claus and Berkerley (1986) and Parry at. al. (1983). Tests performed included both morphological examination and biochemical tests. In the morphological examination, the diameter and length o f cell were measured and the 30 University of Ghana http://ugspace.ug.edu.gh shape and position o f spores noted, The biochemical tests carried out were anaerobic grow th, acid production from D-glucose, hydrolysis o f casein and starch, g row th at pH 5.7, in 6 .5% (w /v) NaCl and 10% (W /V) NaCl, at 37°C and 65°C. The identity o f selected Bacillus isolates were confirmed by testing cultures for the fermentation o f 49 carbohydrates in API 50 CHB galleries (Bio Merieux SA). 3.4.6 Biochemical tests for characterization o f isolates (i) Protein Hydrolysis (Casein utilisation) For decomposition o f casein, plates o f skim milk agar were prepared from lOOg o f skim milk powder and 20g o f agar (Sigma 91 HO 1625, USA) dissolved in 1000ml o f distilled water. The plates o f skim milk agar were each inoculated with a single streak o f the culture incubated at 37°C and examined for clear zones around grow th at 3 to 7 days to indicate decomposition o f casein and proteolytic activity. (ii) Starch Hydrolysis For hydrolysis o f starch, plates o f starch agar were prepared from I Og o f starch and 23g o f nutrient agar (Difco) dissolved in 1000ml distilled w ater were each inoculated in duplicate with the test organism and incubated at 37°C for 3 to 5 days. Hydrolysis o f starch was determ ined at 3 and 5 days by flooding plates with iodine solution Unhydrolysed starch changed colour to blue-black within 15 to 30 min. Hydrolysis o f starch was indicated by clear zones around colonies. (iii) Growth in NaCl To test for grow th in 6.5% sodium chloride (Merck KG, 1.06404, Darmstadt, Germany) tubes containing 3ml o f nutrient broth (Difco 3 6201JA), with 6.5% (W /V) 31 University of Ghana http://ugspace.ug.edu.gh sodium chloride were inoculated w ith the test organism and incubated at 37°C and observed for g row th after 7 and 14 days. This was repeated for g row th in 10% NaCl. (iv) Growth at pH 5.7 To test for g row th at pH 5.7, Sabouraud dextrose agar slants containing lOg o f peptone (Oxoid), 40g o f dextrose and 15g o f agar (sigma) dissolved in 1000ml distilled w ater with pH 5.6 and sabourand and dextrose broth containing lOg o f peptone (Oxoid) and 20g dextrose dissolved in lOOOmi distilled w ater with pH 5.7, were inoculated with isolates which had previously been grown in nutrient broth (Difco). The tubes were observed for grow th for up to 14 days after incubation at 30°C. (v) Acid production To test for acid production from D (+) G lucose, 3ml slants containing 150 (.il o f 10% filter sterile solution o f D (+) Glucose (Merck 8342, Darmstadt) and a basal medium containing lg o f diammonium hydrogen phosphate, Ig o f potassium chloride, 0 2g o f magnesium sulphate (Analytical), 0.2g o f yeast extract, (Difco, 0127-17-9, USA) and 15g o f agar (Sigma, 91 HO 1625, USA), 0 .006g o f bromcresol purple dissolved in 1000ml distilled w ater with pH 7.0, were inoculated with the test organism and incubated at 30°C. Acid production was indicated by a change in colour o f the medium from purple to yellow. (vi) Identification using API 50CHB kit The species o f Bacillus isolates were identified by assaying selected cultures in API 50 CHB galleries (Bio Merieux SA). Isolates were grown in 10ml nutrient broth at 37°C for 48h, plated out on Nutrient Agar and incubated at 37°C for 24h. Bacteria were 32 University of Ghana http://ugspace.ug.edu.gh harvested from the plates with a sterile swab. The cells w ere resuspended in 3ml sterile distilled w ater and its turbidity adjusted to 3 Mcfarland by comparing with a Mcfarland turbidity standard, 150 (il o f the bacterial suspension was transferred into one tube o f 50CHB medium (Bio Mereux SA) and used to innoculate the copul.es o f the API 50 strips containing the dehydrated substrate using a sterile pipette. The strips w ere placed in the boxes containing 10ml o f distilled w ater which has been distributed into the honey comb to maintain moist conditions and incubated at 37(’C. The strips were read at 24h and 48h and the strain identified by referring to a reference table 3.4.7 IDENTIFICATION OF LACTIC ACID BACTERIA (i) Initial identification Gram -positive catalase-negative MRS isolates were examined by gas production from glucose in MRS broth with durham-tube, gas production from MRS broth in which g lucose was replaced w ith g luconate as sole carbon source, grow th at 15°C and 45°C and Hugh and Leifson test (Hugh and Leifson 1953). (ii) Gas Production Production o f gas from glucose was tested for by inoculating MRS broth containing a durham tube with the test organism and incubating at 30°C for at least 72h. Gas production was indicated by an air bubble trapped inside the inverted durham tube. (iii) Growth at different temperatures For grow th at I5°C and 45°C, two tubes containing MRS broth were inoculated with the test organism and incubated at 15°C and 45°C up to 14 days and examined for significant growth. 33 University of Ghana http://ugspace.ug.edu.gh (iv) Hugh & Leifson's test Two tubes each containing 1 litre distilled water, 2g o f peptone (Oxoid L37 England) 5g o f NaOH, 0.3g o f K2HPO4 (Sigma p5504) and 15ml o f 0 .2% bromothymol blue (sigma B 7271) w ere used. 1% glucose (Merck 8342) was added after passing through a sterile filter and the pH o f the solution in the tubes adjusted to pH 7.1. The two tubes were innoculated with the test organism and one o f the tubes topped with paraffin oil to obtain anaerobic conditions. After incubation at 30°C for 2 to 7 days, fermentative reactions were indicated by yellow colouration in the aerobic and anaerobic tubes due to formation o f acid Oxidative reactions were indicated by yellow coloration in the top section o f the aerobic tube whilst anaerobic tubes maintained their original colour o f blue. G ram-positive, catalase-positive cocci w ere tentatively identified by mode o f glucose breakdown, te trad formation, grow th at 15°C and 45°C, grow th at pH 4.2 and 9.3 on 10% (w/v) and 15% (w/v) NaCl agar. Cultures identified as Staphylococcus occurred singly or in pairs, fermented glucose with the production o f CO 2, w ere able to g row at 15°C and 45°C and also mostly in up to 10% NaCl (w/v) but not 15% NaCl (w/v). v) Identification of Lactobacillus species Regular G ram -positive rods and very short rods/coccoids which were catalase negative, oxidase negative, fermentative and grew both aerobically and anaerobically were considered to belong to the genus Lactobacillus, The Lactobacillus species were classified into obligately homofermentative, facultatively heteroferm entative and obligately heterofermentative by their ability to produce (CO2) from glucose and gluconate. 34 University of Ghana http://ugspace.ug.edu.gh 3.5 DETERMINATION OF SUGARS The Lane and Eynon's method was used, The sugar solution was neutralized and clarified. The sugar solution was placed in a 50-ml burette. A preliminary titration was first performed by pipetting 10 or 25 ml o f mixed Fehling's solution into a 300 ml conical flask, 15 ml o f the sugar solution from the burette was added to the mixed Fehling's solution and boiled on an asbestos-covered gauze. Quantities o f the sugar solution w ere added further (1 ml at a time) at 10-15 seconds intervals until the blue colour was nearly discharged. 3-5 drops o f aqueous methylene blue solution (1%) were added and titration continued until the indicator was completely decolourised. An accurate titration was performed by repeating the titration. Almost all the sugar solution required to effect reduction o f the copper was added before heating. It was then boiled gently for tw o minutes and 3-5 drops o f the methylene blue indicator were added. The titration w as completed within a total boiling time o f 3 minutes. The blue colour was fully discharged with the liquid showing an orange-red colour at the end­ point. The proportions o f the various sugars equivalent to 10 or 25 ml o f Fehling's solution were read from a given table. Sucrose was determ ined in a solution which had been inverted and contained no other sugar. The inversion was done on a portion containing about 0 .5g sucrose in a 100 ml volumetric llask with acid and neutralised. This solution was titrated against Fehling’s solution and the amount o f invert sugar produced was obtained by reference to the tables. % Invert sugar x 0.95 = % sucrose 35 University of Ghana http://ugspace.ug.edu.gh 3.6 CHARACTERISATION OF THE PROTEINS OF SOYDAW ADAW A USING GEL ELECTROPHORESIS Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (Excel Gel XL SDS 12-14, Pharmacia Biotech AB 71-7137 00) was used to characterize and compare the proteins in the soydawadawa. 3.6.1 Sample Preparation From a sample o f soydawadawa 0.05g was weighed and added to 500j.il o f Tris bulTcr, pH 7, containing 1.576g Tris/HCl (sigma T-7149), 0 .372g EDTA (Sigma E-5134) per 100ml Milli-Q water. 500mg o f glass beads was added to facilitate disruption o f the cells and the sample w as then placed on ice for Imin and then vortexed for Imin repeating 9 times. The sample was centrifuged at 5 ,000xg for lOmin and the supernatant diluted w ith 90|il o f Tris buffer, pH 8, containing 1.576g Tris/HCL (S igm aT -7149), 0 .372gED TA (Sigma E-5 134) per 1000ml o f milli-Q water. I00f.il o f sample buffer containing 10.0ml Tris buffer, 1.0ml Bromophenol blue solution, 20.0ml Sodium dodecyl sulfate (SDS) solution, 5.0ml Glycerol solution, 1.0ml D ithiothreitol (DTT) solution was added to lOOul o f diluted sample and boiled for 5min. The Bromophenol blue (Merck 8122), contained xg in 1.0ml milli-Q water, the SDS solution 2 .00g o f SDS in x ml o f milli-Q w ater and the Glycerol solution 0 .25g o f Glycerol (M erck 4092) in 5.0ml milli-Q water. A fter boiling, sample was certriluged at 5000xg for 5min. E lectrophoresis was run at x°C for 15min and temperature set to 15°C. About 2-4ml o f kerosene was smeared onto the cooling plate o f the electrophoresis equipment. The 36 University of Ghana http://ugspace.ug.edu.gh gel (Excel Gel XL SDS 12-14, Pharmacia Biotech AB 71-7137-00) was positioned on the cooling plate w ith the cut corner on the gel corresponding to the anodic (+) side of the cooling plate .Buffer strips were positioned on the gel. The sample applicator strip was positioned about 5mm from the cathodic (-) buffer strip and left there during the electrophoresis. Gels w ere stained by silver staining. They were soaked for 30 min in fixing solution containing per 250ml distilled water, 100ml ethanol and 25ml glacial acetic acid. Gels were then placed for 30 min in sensitizing solution containing per 250ml distilled water, 75ml ethanol, 1.25ml o f 25% (w/v) glutardialdehyde, 0.5g sodium thiosulphate and 17g sodium acetate and washed 3 times for 5 min each in distilled water. Gels were then stained for 20 min in a silver solution containing per 250ml distilled water, 25ml o f 2.5% (w/v) silver nitrate solution and 0.1ml o f 37% (w/v) formaldehyde and washed tw ice for 1 min each in distilled water. The stained gels were developed for 2 to 5 min in a developing solution containing per 250ml distilled water, 6 .25g sodium carbonate and 0.05ml o f 37% (w/v) formaldehyde. The developing reaction was stopped after 2 to 5 min by placing the gels in a stop solution containing per 250ml distilled water, 3 .65g EDTA-Na2.2H20 for 10 min and washed 3 times for 5 min each in distilled water. Gels were preserved by soaking for 20 min in preserving solution containing per 250ml distilled water, 75ml o f 87% (w /w ) glycerol and covered with cellophane preserving sheets. 3.7 DETERMINATION OF AROM A COMPOUNDS A GC -MS Hewlett Packard 6890 GC, 5973 MS, Avondale Pennsylvania was used to separate the soydawadawa extract into individual components which were readily identified by the mass spectrometer. 37 University of Ghana http://ugspace.ug.edu.gh 3.7.1 Sample Preparation And Extraction Procedures Soydawadawa sample weighing 2g was mixed with 10ml distilled w ater and 10ml n- hexane in a 100ml separating funnel. The m ixture was shaken vigorously for about a m inute and allowed to separate into aqueous and solvent layers. The solvent layer was filtered through whatman No. 1 filter paper filled with some anhydrous sodium sulphate into a 2ml screw head vial which was then tightly capped. This extract was injected into the GC-MS. 3.7.2 Gas Chromatography And Mass Spectrometry. Aroma compounds present in samples were detected by injecting into a Hewlett Packard 6890 gas chromatography (Hew lett-Packard, Avondale, Pennsylvania) equipped with a fused silica capillary column, (Hp-SMS column) connected to a Hew lett-Packard 5973 mass spectrom eter (Hew lett-Packard, Avondale, Pennsylvania). The column pressure in the Gas Chromatograph column flow was approximately lOml/min and split temperature for the run was 45()C, Ramp and Total Run time, 22min. Analytical grade reagents (NaCl) were used. 3.8 ENZYME ASSAYS Organisms were streaked on casein and starch agar. Those which were more efficient in degrading the casein were selected and their proteinase activities were assayed by using the method described by Young and W ood (1977), Similarly, organism s which 38 University of Ghana http://ugspace.ug.edu.gh were more efficient in degrading starch on agar plates were selected and their a- amylase activities w ere assayed by using the method described by Bernfeld (1955). 3.8.1 Preliminary Screening of cultures for Proteinase and a -amylase activity Cultures w ere streaked on surface dried plates o f casein agar and starch agar (Gordon el. al. 1973), and incubated at 37°C. A fter 3 d plates w ere flooded with iodine solution and the diameters o f the clearing zones were used as an assessment o f proteinase (casein agar) and a-amylase (starch agar) activities. 3.8.2 Preparation o f Extraction Buffer for Proteinase activity The extracting buffer w as 0.1M sodium hydrogen phosphate, pH 6.5. The assay method used was that o f Young and W ood (1977). It has been found useful for analysing proteinases in the presence o f reducing sugars normally found in food substances. Soydawadawa weighing 5g was added to 50ml o f 0.1 M sodium hydrogen phosphate and ground in a m ortar to prepare an extract. The suspension was then washed with 5ml petroleum ether to extract the oil and centifuged at 5000 rpm. The supernatant was stored in a deep-freezer at -20°C. Assays and analyses were carried out on duplicate fermentations and for each sample three determ inations w ere made at each time interval. 3.8.3 Determination o f Proteinase Activity 5ml o f the extract was added to 10ml o f 2% solution o f light soluble casein (BDH) and incubated at 35°C for 30 min. The reaction was term inated by adding 10ml o f 10% 39 University of Ghana http://ugspace.ug.edu.gh trichloroacetic acid (TCA) solution. The m ixture was filtered through whatman No. I filter paper. The optical density o f the filtrate was obtained by reading the absorbance at 275nm with an SP6 250 spectrophotometer. The blank contained the same mixture but with the TCA added simultaneously with the enzyme extract. Enzyme activity was expressed in terms o f an arbitrary unit called an XS unit, and is defined as: 'An enzyme extract which under the stated experimental conditions produced a filtrate with an optical density o f 0 .500 when measured in a 10mm path length cell, had a strength o f 36 XS units per gram ' (Young and W ood, 1977 a). 3.8.4 Preparation of Extraction Buffer for a -Amylase Activity The extracting buffer was 1M potassium hydrogen phosphate, pH 6.5. The assay procedure described by Bernfeld (1955) was used. The extract was prepared as in the proteinase activity with 1M potassium hydrogen phosphate, pH 6.5 as the extracting buffer. 3.8.5 Determination o f a -A inylase Activity 2 ml o f the extract was mixed with 1 ml o f 1% starch solution and incubated for 1 h at 40°C. The reaction was stopped by adding 3ml dinitrosalicylic acid reagent (DNS). The m ixture was heated in a boiling water bath for 5 min, cooled in cold water, and then diluted with 18 ml water. The optical density o f the resultant solution was obtained by reading the absorbance at 550 nm, using an SP6 250 spectrophotometer. The blank was similarly treated except that the DNS was added before adding the starch, solution. The amount o f the reducing sugars formed was calculated from a standard curve prepared with known concentrations o f maltose (Bernfeld, 1955). 40 University of Ghana http://ugspace.ug.edu.gh 3.9 DETERMINATION OF AMINO ACIDS IN SOY DAWADAW A This was done using the P ico-Tag Method, and an Amino Acid Analyzer, Basically an internal standard (norleucine) was added to the protein/sample and hydrolysed with 6M hydrochloric acid. The hydrolysed samples were evacuated to dryness, neutralised with Triethlam ine (TEA ) and sodium acetate. The samples were derivatised with phenylisothiocyanate (P1TC). The derivatives (PTC - AA) were analysed with reverse phase HPLC and UV detection. List and grade o f chemicals used for the am ino acid analysis is shown in Appendix 2. 3.9.1 Working Solutions The solutions used in the analysis were each prepared as follows: 6M hydrochloric acid 1000 ml 37% hydrochloric acid was added to w ater in 2000 ml volumetric flask. M ixture when cold was diluted to mark with water. Internal standard (6.25m M Nor). 0.82g norleucine was weighed, transferred to 1000ml volumetric flask with 6M hydrochloric acid and diluted to mark with hydrochloric acid. It was stored refrigerated. 2.5ni M norleucine 0.328g norleucine was weighed, transferred to 1000ml volumetric flask and dissolved in 17 ml 6M hydrochloric acid. It was diluted to mark with w ater and aliquots stored at - 20°C. 41 University of Ghana http://ugspace.ug.edu.gh 0.1 M Dithiothreitol (DTT) 0.1542g DTT was weighed in 10 ml graduated centrifuge tubes, dissolved in w ater and diluted to mark. (This w as used the same day). 5m M hydroxy proline and taurine 0.655g hydroxyproline and 0.626g taurine, were weighed, transferred to 1000 ml volumetric flask with 17 ml 6M hydrochloric acid and diluted to mark with water. AJiquots w ere stored at - 20"C. Standard 1,1.2, 5m M lm l 2.5mM amino acid standard H and 1 ml 2.5mM norleucine w ere pipetted and vortexed thoroughly. This was stored at - 20°C. Standard 2.1 mM incl. hydroxyproline and taurine 2ml standard 1.1.25 mM and 0.5ml 5mM hydroxyproline and taurine were pipetted, into 4ml sample vials and vortexed thoroughly. This was stored at - 20"C. 0.2 M Sodium acetate 1.7206g water free sodium acetate was weighed, transferred to 100ml volumetric flask and diluted to mark w ith water. This was stored refrigerated. 42 University of Ghana http://ugspace.ug.edu.gh Redry Solution Methanol, 0.2M sodium acetate and Triethylamine (TEA) were mixed in relation 2 :2 :1 in 4 ml sample vials and stored in freezer at (-20°C). Derivatisation Solution Methanol, water, Triethylamine (TEA) and Phenylisothiocyanate (PITC) were mixed in relation 7:1:1:1 and stored at room temperature for 1 h). Opened ampoules o f PITC were divided into 5 x 200 |il in 4ml chromacol tubes, blanketed with nitrogen (30s) and stored at - 20°C up to 3 weeks. Stock EDTA SOLUTION O.lOOg EDTA was weighed, transferred to 100ml volumetric flask with w ater and diluted to mark and stored refrigerated. Eluent A (Acetate buffer), Sample Diluent 38 .Og sodium acetate was weighed in 2000ml beaker and 200 ml water, 1.0 ml TEA and 500 stock EDTA Solution w ere added and mixed thoroughly on a magnetic stirrer. PH was adjusted to 6.1 using Concentrated acetic acid. 1880 ml o f this m ixture was measured in a measuring cylinder and filtered through triton free filter to 2000 ml reservoir bottle. 120 ml acetonitrile was measured and added. This was kept in refrigerated storage. 43 University of Ghana http://ugspace.ug.edu.gh Eluent B (60% acetonitrile) 600ml acetontirile and 400ml water were measured separately in a measuring cylinder and mixed in 2000ml reservoir bottle. 250|il Stock EDTA solution was added, degased for 20 sec and stored refrigerated. 3.9,2 Procedure Sample corresponding to 30 mg protein was weighed into flat bottom flasks. 60 ml 6M hydrochloric acid, 5 ml 6.25mM norleucine and 300 j.iL 0.1 MD TT were added and mixed. Flasks w ere placed in a steel container and placed in a pre heated heater (110HC). Hydrolysis w as done for 22 hours at 1 10°C. Samples were cooled in cold water, transferred to 100 ml volumetric flask, diluted to mark with water, and mixed thoroughly. This was filtered through 0.45 j.un M illipore filter, with 5 ml syringe, into 4 ml chromacol tubes and stored at -20"C. Vacuum work station was connected and started to achieve derivatisation. 2 x 20 (.il standard was pipetted together with 20 (.d o f every sample in 6 x 50 mm tubes. Tubes were placed in reaction vials and these were further placed in the vacuum station. They were evacuated to dryness for 45 min. 30 (j.1 redry solution was added, vortexed thoroughly and evacuated to dryness for 45 min. 20 j.tl derivatisation solution was added and vortexed thoroughly. Samples were left under atmospheric pressure for exactly 10 min and evacuated for 15 min. 40 |il methanol was added, vortexed thoroughly and evacuated to dryness for 2 h. 44 University of Ghana http://ugspace.ug.edu.gh The dried samples w ere dissolved in 200 j.il sample diluent and vortexed thoroughly. The supernatant was transferred into 1 ml syringe. The supernatant was then filtered through Whatman No. 1 filter paper into 4 ml sample vials which were inserted onto the HPLC system. For analysis, samples w ere run on HPLC system controlled by Millenium 2010. Sample set was started with an equilibration time o f 10 min and continued with condition column, run time being equal to gradient time. 10 (.il o f standard and sample w as injected into the column. Eluents were degased and stabilized by continuous helium supply. E luents w ere pre-heated and column heater with tem perature control was used to keep tem perature on column constant. G radient eluting was run, detecting samples by uv absorbance at 254 nm. Peak areas were measured for both the standard and samples and corrected by internal standard. 45 University of Ghana http://ugspace.ug.edu.gh 4.0 RESULTS AND DISCUSSION 4 1 FIELD SURVEY A brief field study confirmed that different types o f raw materials are fermented into dawadawa and that the preferred raw material for preparing dawadawa varies from one ethnic group to the other. 60% o f the dawadawa found on the m arkets were made from African locust beans, w ith 25% from Soybean, 10% from a m ixture o f African locust beans and soybean mixture and 5% from a mixture o f African locust beans and groundnuts (Arachis hypogaea). The processing method used by all the ethnic groups is the same and the dried products are stored in traditional earthernware pots, baskets or boxes. They are used almost daily in cooking. About 8g to 12g at a time is crumbled into a pot o f soup or stew as a seasoning. This is enough for a meal for about 4 people. Several w orkers including Ogbadu and Okagbue (1988) have reported that both African locust beans and soybeans are used as raw material by traditional dawadawa processors. As to how much dawadawa is often taken per person per day, the consumers gave am ounts that ranged between 2g to 4g respectively per person per day. According to consumers interviewed, dawadawa could be roasted in whole and used like a piece o f fish in eating kenkey, gari o r any other food Simmons (1976) found that the average daily per capita intake o f "dawadawa" among some Hausas o f Northern Nigeria constitutes 1,4% o f the daily calories intake and 5% o f the total protein intake. Lawson (1965) found the average per capita per day consumption o f dawadawa in Togo and Ghana to be 4 and 2g respectively. On the other hand, Dema (1965) found that the Yorubas o f South Western Nigeria consume lOg per day per 46 University of Ghana http://ugspace.ug.edu.gh person, whilst the overall consumption estimated for parts o f Nigeria by Nicol (1959) range from 1 to 17g per person per day. Ogunbunmi and Bassir (1980) reported that dawadawa contributes appreciable amounts o f protein to the diet o f Nigerians. The results o f the survey showed that the preparation o f the dawadawa depends on the type o f raw material used. To make soybean dawadawa, the soybeans are first roasted for about 30 minutes to a brown colour and pounded to remove the seed coat o r testa. The dehulled soybeans are cooked in w ater for 1 h, drained using a sieve and spread in a basket lined with leaves after cooling. The basket is covered with more leaves and placed in a warm place for 2 to 3 d for fermentation to take place. The fermented beans are moulded into various shapes, sun dried and stored as dawadawa in pots, baskets, etc. (Figure 3.1) In the case o f the African locust bean, the beans are boiled for 24h to soften the seed coat o r testa. The boiled seeds are put in a mortar and pounded slightly with a pestle o r pressed by foot to remove the softened testa. The seeds are rubbed between the palms or against the walls o f a basket. The cotyledons are then washed thoroughly and the testa removed. The washed bean cotyledons are boiled again for 1 to 2 h in a metallic pot and the hot bean cotyledons drained through sieves or baskets. They are spread in a basket lined w ith leaves after cooling and covered with more leaves. Millet flour or wood ash is sprinkled on the bean cotyledons before fermenting for 72 h. A fter fermentation, they are moulded into various shapes, sun dried for a day or two and stored as dawadawa in pots, baskets etc. (Figure 2-1). In the case where groundnuts are added to either the African locust bean or soybean to prepare the dawadawa, the groundnuts are given the same treatment as the soybean. 47 University of Ghana http://ugspace.ug.edu.gh The main differences found between the method o f production o f the various types of dawadawa were mainly in time and energy used. The method o f production ot soybean dawadawa saves time, conserves energy and thus lead to increased production. In term s o f product acceptance, consumers rated dawadawa processed from the African locust beans as the best product. This was followed by dawadawa processed from soybeans and lastly dawadawa processed from a m ixture o f African locust beans and soybean. On the other hand, dawadawa prepared from a m ixture o f African locust beans and groundnuts is not very well patronised. According to consumers interviewed, a good quality dawadawa has a strong ammonia-like smell and a dark brown colour. After fermentation, the seeds are either moulded in seed form or pounded into a paste before moulding. Figures 1 and 2 indicate boiled and roasted soydawadawa. 48 University of Ghana http://ugspace.ug.edu.gh III !j FIG 4-1 BOILED SO YDAW ADAW A 49' University of Ghana http://ugspace.ug.edu.gh FIG 4-1 BOILED SO YDAW ADAW A •49' University of Ghana http://ugspace.ug.edu.gh FIG 4-2 RO ASTED SO YDAW A DAW A 50 University of Ghana http://ugspace.ug.edu.gh 4.2 CHEM ICAL COM POSITION OF SOYDAW ADAW A The proximate composition o f soydawadawa is shown in Table 4-1. The percentage moisture o f the soydawadawa was generally higher than the soybeans. However, as expected, the roasted product had a slightly lower moisture content than the boiled product because the boiled product absorbed w ater during boiling. W ork by Barimalaa el al. (1994) showed an increase in moisture, protein and fat contents o f cotyledons during fermentation o f soybeans into dawadawa. However, this study though confirm ing an increase in protein content showed a decrease in the fat content o f the soydawadawa during fermentation. Obizoba and Atu (1993) in their work on the chemical evaluation o f some food condiments o f Nigeria also confirmed an increase in the protein content o f African locust bean dawadawa during fermentation. They observed that the 4-day fermentation period caused the highest increases in protein o f the dawadawa and stated that fermentation for 4-days offers a greater advantage than o ther periods for production o f nutritious and cheap food condiments in Nigeria. The increase in protein content during fermentation o f dawadawa was explained by Lee el. al. (1983) in their work on the nutritional evaluation o f naturally fermented soybean and the enzymatic activity changes during the preparation. They found that, the activities o f protease and lipase increased during fermentation with progress o f proteolysis and a release o f more free amino acids. Heat has been known to cause an improvement in some proteins particularly in Legumes. The enhancement o f the nutritive value o f soybean protein by moderate heat is generally attributed to the destruction of the heat-labile antitryptic factor naturally present in the raw bean. Dry heating is less effective than steam in enhancing nutritional value and the presence o f 51 University of Ghana http://ugspace.ug.edu.gh water partially prevents the adverse effect o f excessive heating (Harris and Von Loesecke, 1960). This may explain why the protein content o f the boiled soydawadawa was higher than that o f the roasted soydawadawa. The chemical compostion o f dried soybean seeds was given by Carrao el. al. (1994) as w ater - 5.0 - 9.4% , protein - 29.6 - 50.3% , Fat - 13.5 - 24.2% and Ash - 3.3 - 6.4% . The chemical composition o f the dried soybean seeds found in this study falls within the ranges given by Carrao cl. al. (1994) for yellow seeded cultivars o f soybeans used in this study. Table 4-1 Proximate chem ical composition o f Soydawadawa % Dry M atter Basis % Soybean % Soydawadawa (BOILED) % Soydawadawa (ROASTED) M oisture 8.8 25.0 24.0 Ash 5.2 3.5 5.0 Protein 42.8 45.5 43.2 Fat 36.7 26.7 21.5 * Average values for 2 sets o f samples 4.3 CHANGES IN PH AND TITRATABLE ACIDITY DURING FERM ENTATION The pH o f soybeans changed from acid thorough neutral to alkaline during fermentation (Table 4-2). A progressive increase in the titratable acidity was recorded from the beginning to the end o f the fermentation period. Changes in the pH and titrable acidity during fermentation are shown in Table 4.2. The pH o f boiled soybeans increased from about 6.6 to 8.3 and roasted soybeans from about 6.4 to 8.2 during 72 h. Titratable acidity o f boiled soybeans increased from about 0.1 to 0.4 and roasted 52 University of Ghana http://ugspace.ug.edu.gh soybeans from about 0.1 to 0.4 during 72 h. The simultaneous increase in pH and acidity is unusual since one would expect the pH to drop as acids are produced. Hesseltine (1965), suggested that the simultaneous increase in pH and acidity during fermentation o f legumes may be due to the high buffering capacity o f the legume beans and the proteolytic activities o f Bacillus subtilis leading to ammonia release. This phenomenon is characteristic o f most vegetable protein fermentations. This observation was also made by Odunfa (1985), when he found that proteolysis was the major biochemical change in fermenting African locust bean during iru fermentation. W agenknecht cl. al., (1961), made similar observations about tempeh fermentation. The pH o f the fermenting soybeans increased despite the large amount o f acid that was liberated. They suggested that liberated ammonia or other basic end products o f protein decomposition was the cause o f this. The final product had comparatively lower pH and titratable acidity values. This could be due to the fact that during the subsequent drying o f the product after 72 h o f fermentation to obtain the final product, there was loss o f w ater leading to reduced w ater activity as well as reduced microbial activity which as a result reduced the release o f ammonia as well as the pH o f the fermenting medium. 53 University of Ghana http://ugspace.ug.edu.gh Table 4-2: Changes in pH and Titratable acidity (expressed as lactic acid) during the fermentation o f soybeans into dawadawa. FERM ENTATION BOILED ROASTED TIME SOYDAW ADAW A SOYDAW ADAW A PH Acidity pH Acidity 0 h 6.59 0.07 6.41 0.12 24 h 6.87 0.39 6.57 0.42 48 h 8.00 0,35 7.90 0.51 72 h 8.25 0,42 8.15 0.42 * Final Product 6,38 0.02 6.1 1 0.02 '' Average values for 2 sets o f samples * Soydawadawa dried after 72h o f fermentation. 4.4 M ICROBIAL POPULATION OF FERM ENTING SOYDAW ADAW A The viable counts obtained on the soydawadawa are shown in Table 4-3. The counts obtained for the MRS Agar plates were considerably lower than those on the PCA Agar plates. The low level o f m icro-organisms at the start o f fermentation is likely to have resulted from the method o f processing which involved heating o f the beans. It is possible that this served as a selective mechanism for the fermenting m icro-organisms At 37"C, the processed beans (i.e. at 0 h) count on PCA was about 10 ’ cfu/g for both the boiled and roasted soydawadawa whereas that o f fermented beans (72 h) was about 10n cfu/g and 109 cfu/g. 54 University of Ghana http://ugspace.ug.edu.gh 4 .4 .1 Aerobic M esophilic count In all fermenting soydawadawa samples, the population o f aerobic mesophiles enumerated on PCA consisted o f a variety o f G ram-positive catalase positive rods bearing phase bright spores as well as Gram-negative bacteria and Gram-positive catalase-negative rods at the start o f fermentation (Table 4-3). The population o f aerobic mesophiles o f all fermented products after 48 h fermentation w ere dominated by the Gram positive catalase positive rods. In most fermenting samples, the Gram- positive catalase negative rods and coccobacilli found at the start o f fermentation were greatly reduced in the flora o f the final fermented product. The G ram -negative bacteria present at the start o f the fermentation were not found in the flora o f the final fermented product. The Gram-positive, catalase-positive rods bearing phase bright spores which grew on PCA and were subcultured in nutrient broth or agar, were considered to be Bacillus species. The G ram -positive catalase-negative rods were similar in colony and cell m orphology to colonies which w ere later found to dom inate grow th on de Man Rogosa Sharpe (MRS) Agar plates. The Bacillus species were often present at moderately high levels, about I0'f cfu/g. Some G ram -negative bacteria were found in all two types o f samples but were usually present at levels o f less than 10 3 cfu/g. 4.4.2 Growth On de M an Rogosa Sharpe (M RS) Medium The initial population o f soydawadawa enumerated anaerobically on MRS agar consisted ot G ram-positive catalase-negative rods and coccobacilli with rods as the 55 University of Ghana http://ugspace.ug.edu.gh dominant type. Representative colonies o f the G ram -positive catalase-negative rods and coccobacilli were found to be oxidase-negative non-sporing and strictly fermentative. They were therefore classified as lactic acid bacteria. The G ram -positive catalase-negative rods and coccobacilli found on PCA plates were also determ ined to be lactic acid bacteria. The population o f the fermenting soydawadawa samples on MRS sometimes contained several relatively large colonies which were determ ined to be yeasts by morphological examination. The population o f lactic acid bacteria increased from a level o f 10 3 to 10 6 cfu/g and 10 4 to 10 6 cfu/g for boiled soydawadawa and roasted soydawadawa respectively during fermentation (Table 4-3). 56 University of Ghana http://ugspace.ug.edu.gh Table 4-3: M icrobial population o f Bacteria and Yeast o f fermenting Soydawadawa (cfu/g). FERMENTATION TIME BOILED SOYDAWADAWA Count (cfu/g) ROASTED SOYDAWADAWA Count (cfu/g) Start o f Fermentation Aerobic mesophiles 0 2.5 x 103 9.0 x 10’ Lactic acid bacteria (b) 6.1 x 10' 6.3 x I04 Yeasts (c) 1.5 x 103 2.0 x 10 After 24 hours Aerobic mesophiles 4.6 x 109 1.2 x 10° Lactic acid bacteria 2.5 x 105 5.0 x I05 Yeasts 2.7 x 10s 1.0 x 10 After 48 hours Aerobic mesophiles 1.2 x 109 9.0 x 109 Lactic acid bacteria 1.9 x 106 7.5 x 10f> Yeasts 3.1 x 102 4.6 x I0 ; After 72 hours Aerobic mesophiles 1.5 x 10" 1.8 x I09 Lactic acid bacteria 1.4 x 106 5.9 x I05 Yeasts No grow th No grow th Final Product Aerobic mesophiles 6.2 x 108 5.1 x 10s Lactic acid bacteria 5.4 x 10* 2.2 x I04 Yeasts No grow th No grow th a. Counts on plate count Agar including both G ram-positive and Gram-negative bacteria. b. Enumerated on MRS Agar reflecting Gram-positive, catalase-negative rods and coccobacilli. c. Determined on Malt Agar. 57 University of Ghana http://ugspace.ug.edu.gh Table 4-4: Morphological and biochemical characteristics o f Bacillus isolates ISOLATE TEST 1 2 3 4 5 6 7 8 9 Cell D iam eter > 1.0 - + + - - Spore Shape C O/E E E E E 0 c c Spore position T T/C C T/C C T/C C c c Gram stain + + + + + + + + + Catalase reaction + + + + + + + + + Oxidase reaction + + + + + + + + + Anaerobic grow th - V + - - V + + Acid from D -G lucose + + + + + + V + + Acid from L-arabinose + - + + + + + + Acid from D-Mannitol + - + + + + + -I- + Gas from G lucose - Casein hydrolysis + + + + + + + + + Starch hydrolysis + + + + + + + + Grow th in 6.5% NaCl + + + + + + + + + Grow th at 65°C - - - - G row th at 30"C + + + + + + + t t Specics o f Bacillus identified /. B. subtiHs C - Central / Circular 2. B. cere us E- Elongated / Ellipsoidal 3. B. licheniformis 0 - Oval 4. B. sub til is T- Terminal 5. B. subli/is 6. B. firm us 7. B. subli/is (S', B. subli/is B. subli/is 58 University of Ghana http://ugspace.ug.edu.gh The viable counts obtained for the malt agar plates were very low (Table 4-3). By morphological examination, the large colonies on the malt agar plates w ere determ ined to be yeast. However, after 72 h o f fermentation, there was no grow th. Further investigations were not carried out to identify the exact species o f yeast cells lkenebomeh (1982) and Campbell-Platt (1980) have reported the incidence o f a few fungi in the dawadawa fermentation as contaminants. Thus the incidence o f a few yeast cells in the soydawadawa in this study could be as a result o f contam ination. The bacteria ou tgrew the fungi and possibly created an environment that was not conducive for the grow th o f fungi. 4 5 PATTERN OF M ICROBIAL GROW TH Microbial counts increased from the start o f fermentation to the end o f fermentation after 72 h (Table 4-2). However, there was reduction in microbial counts for the final product. This reduction in microbial counts for the final product could be due to reduced w ater activity in the final product after drying which possibly in combination with other factors, resulted in depressed microbial grow th and biochemical activities in general. A lthough microbial counts generally increased with time o f fermentation, the rate o f increase o f aerobic mesophiles was higher than that o f the lactic acid bacteria due to increased pH in the sample. High pH normally promotes grow th of' aerobic mesophiles like Bacillus species but is not very conducive for grow th o f lactic acid bacteria. 4.4.3 Yeast and Moulds Work by lkenebomeh (1989) on the influence o f salt and temperature on the natural fermentation of African locust beans showed that there was no microbial grow th on 59 University of Ghana http://ugspace.ug.edu.gh acidified PDA known to select for yeasts and moulds. This indicates that moulds and yeasts are not involved in the fermentation o f the African locust bean seeds. However, in the present work, microbial grow th on Malt Agar indicated the presence o f yeasts, though they may not play much role in the fermentation process. W ith increased pH during fermentation, the yeast disappeared in the latter days o f fermentation. 4.6 OCCURRENCE OF DIFFERENT BACILLUS SPECIES The most frequently isolated cultures formed colonies with irregular margins and rough ridged or ringed surfaces often producing exudate. They had small cells with circular centrally placed spores, and were identified as strains o f Bacillus suhtilis. The morphological and biochemical tests showed that most o f the Bacillus isolates from the fermenting soydawadawa produced acid from D-glucose, L-arabinose, D- xylose and D-mannitol. They also hydrolysed casein and starch, reduced nitrate, grew at pH 5.7 and 6.8 and in 6.5% NaCl, (Table 4.4). These Bacillus subiilis strains generally utilized ribose, a-m ethyl-D -glucoside, amygdalin, esculin, salicin, L- arabinose, cellobiose, maltose, saccharose, trehalose, inuline, D-raffinose, amidon, glycogen, glycerol, D-xylose, D-glucose, D -fructose, D-mannose, inositiol, mannitol, sorbitol, arbutin, D -turanose, melibiose and gentiobiose, in API 50 CH galleries (Table 4-5). 60 University of Ghana http://ugspace.ug.edu.gh Table 4-5; Percentage o f identified species from boiled and roasted soydawadawa ________ which fermented various carbohydrates___________________ Carbohydrate B. subtilis Ii. cereus li. licheniformh Ii. pumilus Ii. firmus Glvccrol r ioo 0 100 100 100 Ervthritol 0 0 0 0 0 D-arabinose 0 0 0 0 0 L- arabinose 100 100 100 100 0 Ribosc 100 0 100 100 0 D-xvlosc 83 0 100 100 0 L-xvlosc 0 0 0 0 0 Adonitol 0 0 0 0 0 P-mcthvl-wlosidc 0 0 0 0 0 Galactose 0 0 0 100 0 D-glucose 100 100 100 . 100 100 D-fructosc 100 100 100 100 too D-mannose 100 0 100 100 0 L-sorbosc 0 0 0 0 0 Rhanmosc 0 0 0 0 0 Dulcitol 0 0 100 0 0 Inositol 100 0 100 100 0 Mannitol 100 0 100 100 100 Sorbitol 100 0 0 100 0 a-methvl-D-mannoside 0 0 100 100 0 a-methvl-D-glucoside 100 0 100 100 0 N acetvl Rlucosaniinc 33 100 100 100 100 Amygdaline 83 0 100 100 0 Arbutin 100 0 100 100 0 Esculin 100 100 100 100 100 Salicin 100 0 100 100 0 Cellobiose 100 0 100 100 0 Maltose 100 100 100 100 100 Lactose 17 0 0 0 100 Mclibiose 100 0 0 0 0 Saccharose 100 100 0 0 0 T rchalose 100 100 100 0 100 Inulin 100 0 0 0 100 Mclezitose 0 0 0 0 0 D-raffinose 100 0 0 100 0 Amdion 100 100 100 100 0 GIvcorcii 100 100 100 0 100 Xvlitol 0 0 0 0 0 p-L>cntiobiose 17 0 100 0 0 D-turanose 83 0 0 100 0 D-lsxose 0 0 0 0 0 D-tagaiosc 0 0 0 0 0 L-fucose 0 0 0 0 0 D-arabitol 0 0 0 0 0 L-arabitol 0 0 0 0 0 Gluconate 100 0 0 0 0 2 Ccto-i>luconate 0 0 0 0 0 5-Ceto-gluconatc 0 0 0 0 0 61 University of Ghana http://ugspace.ug.edu.gh Within the Bacillus subli/is isolates, variations could be observed in the colony morphologies pointing to the possible presence o f different strains of the species. O ther Bacillus species identified were Bacillus pumilus, Bacillus liclieiiifonnis, Bacillus cereits and Bacillusfirmus. Isolates identified as Bacillus lichenniformis had opaque rough star shape colonies which were very strongly attached to the surface o f nutrient agar and difficult to scrape off. They could grow anaerobically. Isolates identified as Bacillus cereus generally had oval colonies with pear shaped ends with a medium to coarse matt appearance. Isolates identified as Bacillus punrUus had small creamy coloured colonies with moist raised surfaces. Variations were observed in the proportions o f the different Bacillus species in the two sets o f samples examined but Bacillus subli/is accounted for over half o f the Bacillus population in the two samples. At all stages o f the fermentation, the dominant species was Bacillus sublilis (Table 4- Table 4-6: D istribution (percentage) o f different Bacillus species in Soydawadawa. (Final Product). % BACILLUS SPP % OF ORGANISM BOILED SOYDAWADAWA ROASTED SOYDAWADAWA B. sublilis 50 48 B. pumilus 20 20 B, Ucheniformis 6 7 B. cereus 16 16 B. firmus 8 9 * A total o f 224 isolates from two samples o f soydawadawa taken at a 24 hour interval for 3 days. 62 University of Ghana http://ugspace.ug.edu.gh The characteristics reported for the dominant bacterial isolates are similar to those of Bacillus subtilis described by Gordon (1973). Odunfa (1981) reported the presence o f Bacillus subtilis in fermenting African locust beans, while Bacillus species amongst o thers were also reported to be involved in the fermentation o f dawadawa (lkenebomeh t>t at. \ 1981; Antai and Ibrahim, 1986). 4.7 THE ROLE OF BACILLUS SPECIES IN SOYDAW ADAW A FERM ENTATION Fairly severe heat pre-treatm ents are given to the soybeans during the preparation o f the Boiled and Roasted Soydawadawa. This pretreatment could be described as a spore activation process favouring selection for Bacillus species by virtue o f their heat resistant spores. The findings that Bacillus species play a dominant role in the fermentation o f a traditional staple product is unexpected due to the reported clinical importance o f the species (Parry et. al. 1983). Since the major constituents o f soybeans are proteins, fats and carbohydrates, the organisms responsible for fermenting them must be capable o f utilizing these three constituents. M ost o f the organisms isolated from the fermented beans are known to possess such characteristics. Bacillus species have a proteolytic ability and are also able to break down oils (Frazier 1967; Forgarty and Griffin 1973). The results o f this work demonstrating the central role o f Bacillus subtilis in the fermentation o f soybeans into soydawadawa is in agreement with the findings o f other investigators who have studied the fermentation o f soybeans in other countries. 63 University of Ghana http://ugspace.ug.edu.gh Tamang (1993) and Sakar el. a l (1994) in their work on Kinema, an indigenous non­ salted fermented soybean food o f the Himalayan regions showed the presence ot Bacillus sublilis, Enterococcus faecium , Candida parapsilosis and Geotrichum candidum. O f the organisms isolated from Kinema, B. sublilis was the only one found to play a role in the fermentation (Sakar and Tamang, 1994). Ogbadu and Okagbue (1988) in their work on the bacterial fermentation o f African locust bean for dawadawa production also found Bacillus sublilis and Bacillus pumilus to be the dominant organisms in the fermenting locust beans with Bacillus sublilis dom inating the fermentation process. The findings o f the present work are similar to those reported for African locust bean dawadawa. Campbell-PIatt (1980) as reported by Odunfa (1985) found about 3 1% o f m icroorganisms isolated from numerous dawadawa samples collected from different countries to be Bacillus sublilis. In some samples Campbell-PIatt (1980) found Bacillus sublilis to constitute 61-69% o f all isolates. Bacillus sublilis has also been confirmed as the predominant species in dawadawa fermentation in Nigeria by Odunfa (1981), Ikenebomeh (1989), Adewuyi (1983) and Odunfa and Oyewole (19S6). 4.8 CHARACTERIZATION OF THE LACTIC ACID BACTERIA POPULATION Over half o f all G ram-positive catalase-negative rods and coccobacilli enumerated on MRS from fermenting soydawadawa samples and previously determ ined to be lactic acid bacteria were found to be facultatively heterofermentative lactobacilli. This was done by examination o f their ability to produce C 0 2 from glucose and gluconate (Table 4-7). There were also some homofermentative Lactobacilli. The / .actobaciUus 64 University of Ghana http://ugspace.ug.edu.gh species were non-oxidative, metabolized glucose fermentatively in Hugh and Leilson medium, grew at pH 4.7 and 9.6 but not in 6.5% NaCl and 18% NaCl, I he species of the lactobacil/i w ere not identified. Table 4-7 Biochemical characteristics o f Lactic acid bacteria isolated from Boiled and Roasted Soydawadawa TEST 1 2 3 4 5 6 Tetrad formation - - - - Gram stain + + + + + + Oxidase test - - - Catalase test + - - Anaerobic grow th + + + + + + C 0 2 from glucose - + + + Grow th at 45° C + V Grow th in 6.5% NaCl + + + - + + Grow th at pH 4.4 + + + V + + Fermentative (H & L) + + + V + + Oxidative (H & L) V absent + present V Variable 65 University of Ghana http://ugspace.ug.edu.gh 4.9 TOTAL REDUCING SUGARS The sugar levels showed a remarkable fluctuation with the length o f fermentation. I he reducing sugar level increased during the first 24h but subsequently decreased (Figure 4-3) This trend in sugar levels during fermentation has also been described by Odunfa (1985) in his studies on the biochemical changes in fermenting African locust bean (Parkin big/obosa) during ‘iru’ fermentation. He explained that, the initial rise in the reducing sugar level may not be due to the amylase activity. They might be produced from the hydrolysis o f oligosaccharides, present in the unfermented bean (Odunfa, 1983). These sugars are easily untilisable by the Bacillus species involved in the fermentation. A lthough carbohydrates constitute 11-15% o f the unfermented beans no starch has been reported in the fermented beans (Watson, 1971). Eka (1980) also recorded a decrease in the carbohydrate level o f unfermented locust beans and fermented beans. However a similar work by Odunfa (1982) during “ogiri”, a fermented melon (citrullus vulgaris) product showed an initial decrease in the sugar level. He explained that the initial decrease in the sugar level may be due to the initial population o f bacteria which preferentially utilise tlie soluble sugars in the melon (Odusote, 1977) found that melon contains 2.5% soluble sugar and I 1% starch by weight. After the exhaustion o f the sugars, the initial population is succeeded by amylolytic bacteria which hydrolyse the starch, thereby increasing the sugar level in the fermenting melon. O dusote (1980) found that the latter stage o f “ogiri” fermentation was charcterised by a - amylase producing Bacillus species. 66 University of Ghana http://ugspace.ug.edu.gh Total reducing sugars determ ination Fig 4-3 Total sugars of: BSD BSD BSD BSD BSD RSD RSD RSD RSD RSD (0h)= 80 (24h)= 130 (48h)= 65 (72h)= 15 (final product)= 5 (0h)= 78 (24h)= 125 (48h)= 68 (72h)= 12 (final p roduct)= 5 67 University of Ghana http://ugspace.ug.edu.gh University of Ghana http://ugspace.ug.edu.gh 4.10 ENZYM ATIC ACTIVITY IN SOYDAW ADAW A DURING FERM ENTATION All isolates o f Bacillus species, the dom inating m icroorganisms showed a similar pattern o f enzymatic activity with respect to a-am ylase and proteinase activity. All Bacillus isolates showed high proteinase activity and nearly all except Bacillus puwilus showed high amylase activity. 68 University of Ghana http://ugspace.ug.edu.gh Diameter o f clear zone on skim milk agar ind icating extent o f proteinase activity o f m icro-organism . Fig. 4.4 Microorganism BS 1 = Bacillus subtilis 1 BS 2 = Bacillus subtilis 2 BS 3 = Bacillus subtilis 3 BS 4 = Bacillus subtilis 4 BS 5 = Bacillus subtilis 5 BS 6 = Bacillus subtilis 6 BC = Bacillus cereus BP = Bacillus pumilus BL = Bacillus licheniformis BF = Bacillus firmus 69 University of Ghana http://ugspace.ug.edu.gh BS1 BS2 BS3 BS4 BS5 BS6 BC BP BL BF Micro-organism Fiq.^.^Piameter of clear zone on skim milk agar indicating extent of proteinase activity of micro-organisms University of Ghana http://ugspace.ug.edu.gh Fig. 4.5 Proteolytic activity in soydawadawa during ferm entation. BSD - Boiled soydawadawa RSD - Roasted soydawadawa Proteinase activity o f : BSD (0h)= 0.18 BSD (24h)= 1.00 BSD (48h)= 3.00 BSD (72h)= 4.10 BSD (final product)= 3.50 RSD (0h)= 0.30 RSD (24h)= 1.10 RSD (48h)= 2.50 RSD (72h)= 3.30 RSD (final p roduct)= 2.50 70 University of Ghana http://ugspace.ug.edu.gh ■BOILED -ROASTED FERMENTATION TIME (H) FIG. 4-3 PROTEOLYTIC ACTIVITY IN SOYDAWADAWA DURING FERMENTATION. University of Ghana http://ugspace.ug.edu.gh The proteinase activity o f the Bacillus isolates was determ ined by their ability to break down skim milk. All the Bacillus isolates had proteinase activity which was exhibited by their ability to disintegrate skim milk when directly placed onto sterile skim milk agar plates. All the strains o f Bacillus si/blilis exhibited a higher level o f proteinase activity than the other Bacillus species. Bacillus Ucheniformis isolates examined showed very low proteinase activity and w ere in fact the least proteolytic with a clearing zone o f about 12 mm in diameter (Figure 4.4). The results indicated that strains within the species group differed in the quantity o f extracellular proteinase secreted. Also, the proteinase activity increased w ith increase in time o f fermentation. The most significant biochemical change that occurs during dawadawa fermentation is protein hydrolysis (Odunfa, 1985). High proteinase activity gives rise to rapid amino acid production. The proteinase activity within the fermenting soydawadawa was also determ ined by measuring the level o f enzyme activity at specific times during the period o f fermentation. The proteinase activity in the boiled and roasted soydawadawa medium is shown in Fig, 4-5, 4.10.1 Proteinase Activity 71 University of Ghana http://ugspace.ug.edu.gh At the start o f fermentation, there was detectable proteinase activity despite the fact that there w ere mostly spores present. The proteinase activity in the Bacillus isolates decreased in the order o f BSI = BS3 = BS4>BS6 = BF>BS5>BS2>BC>BP>BL as shown in Fig. 4.4. In the few hours o f fermentation, there w as a rapid increase in proteinase activity w ith the activity reaching a maximum level between 60 and 72h. This was almost followed by an abrupt proteinase activity decline at the end o f the fermentation period when the product w as dried. The above observation falls within expectation. Normally bacteria have a peculiar pattern o f growth. At the log phase or initial stage where no multiplication takes place little enzyme activity is observed. This phase is followed by the logarithm ic phase where bacteria numbers approximately double w ith each interval tim e to com e to a stationary and decline phase where no multiplication occurs and numbers decrease. In industrial production o f many important secondary metabolites, p roduct formation generally occurs in the idiophase, i.e. after the initial period o f rapid grow th as reported for penicillin. Y okotsuka (1972) and Yamamoto et al. (1972) also concluded that around 60 h is an optimal cultivation time for enzyme production. The results obtained from Fig. 4-5 were subjected to statistical analysis using two-way Analysis o f Variance (ANOVA) and Duncan Multiple Range Test (DM RT). The levels o f significance chosen was at a = p(0.01) (Appendix 3). B o th fermentation time and method o f production significantly affected proteinase activity (Table 4-9). P roteinase activity increased w ith fermentation time. The proteinase activity o f the final product was lower and this could be attributed to the reduced m icrobial load o f the final product. The boiled soydawdawa showed higher proteinase activity than the roasted soydawadawa as shown in Table 4-9. However the interaction between the 72 University of Ghana http://ugspace.ug.edu.gh type o f soydawadawa and time o f fermentation did not have any significant effect on proteinase activity. Table 4-9. P roteinase activity was affected by only fermentation time and type o f soydawadawa. Table 4-8 Mean effect o f fermentation tim e and type o f Soydawadawa on the Proteinase activity o f Bacillus Species Fermentation Proteinase activities in (units) T ime/h Boiled Roasted Soydawadawa Soydawadawa 0 0.20 0.30 24 1.00 1.10 48 3.00 2.45 72 4.10 3.60 Final product 3.50 2.45 Proteinase Activity M ean * 2.36 1.98 a = 0.05 4.10.2 a - Amylase Activity Bacillus pumilus did not show any a-amylase activity irrespective o f prolonged incubation. However, the other Bacillus isolates showed high a-am ylase activity. The production o f amylolytic enzymes by representatives o f the genus Bacillus is common but not general. Species such as Bacillus subtilis, Bacillus licheniformis, Bacillus megaterium, Bacillus circulans, Bacillus maecerans, Bacillus coagulans, Bacillus stearothermophilus and Bacillus brevis produce a-amylase. (Zemek et. a l , 1980). 73 University of Ghana http://ugspace.ug.edu.gh The level o f activity in the Bacillus isolates decreased in the o rder o f I3S2 - BS5>BS3>BS4 = BL>BS6>BC>BF>BP (Fig. 4.6). However the interaction between the type o f soydawadawa and time o f fermentation did not have any significant effect on a-am ylase activity. Table 4-10. a-am ylase activity was affected by only fermentation time and type o f soydawadawa. Boiled soybeans processed into soydawadawa did not show any detectable a-am ylase production at the start o f fermentation as shown by the roasted soydawadawa in Fig. 4.7. The trend in production o f the a-am ylase was similar to the production o f proteinase except that after 48 h both the boiled and roasted soydawadawa showed similar levels and almost overlapped. Only fermentation time significantly affected a - amylase activity (Table 4-10). 74 University of Ghana http://ugspace.ug.edu.gh Figure 4.6 D iam eter o f c lear zone on starch agar ind icating extent o f a -am y la se activity o f m icro - organisms Microorganism BS 1 = Bacillus subtilis 1 BS 2 = Bacillus subtilis 2 BS 3 = Bacillus subtilis 3 BS 4 = Bacillus subtilis 4 BS 5 = Bacillus subtilis 5 BS 6 = Bacillus subtilis 6 BC = Bacillus cereus BP = Bacillus pumilus BL = Bacillus licheniformis BF = Bacillus firmus 75 University of Ghana http://ugspace.ug.edu.gh D ia m et er o f cle ar zo ne (m m ) BS1 BS2 BS3 BS4 BS5 BS6 BC BP BL BF Micro-organism Fig. 46Diameter of clear, zone on starch agar indicating extent of rv -amylase activity of micro-organisms University of Ghana http://ugspace.ug.edu.gh Fig 4.7 a - Amylase activity in soydawadawa during fermentation. a - Amylase activity o f Bacillus species BSD (0h)= 0.00 BSD (24h)= 0,20 BSD (48h)= 0.30 BSD (72h)= 0.80 BSD (final product)= 0.65 RSD (0h)= 0.10 R SD (24h)= 0.25 RSD (48h)= 0.32 RSD (72h)= 0.82 RSD (final product)= 0.68 76 University of Ghana http://ugspace.ug.edu.gh -A M YL AS E AC TI VI TY (M G. R ED . SU G 7M L. F IL TE R) University of Ghana http://ugspace.ug.edu.gh Table 4-9 M ean effect o f fermentation tim e and type o f Soydawadawa on a - Amylase activity o f Bacillus Species Fermentation a-A m ylase A ctivity in (units) Time/h Boiled Roasted Soydawdawa Soydawadawa 0 0.00 0.10 24 0.20 0.25 48 0.30 0.32 72 0.80 0.82 Final product 0.65 0.68 a-Am ylase Activity M ean 0.39 0.43 a = 0.05 4.11 AM INO ACID PROFILE OF SOYDAW ADAW A The amino acid profile o f soybeans did not change much w ith fermentation (Table 4- 11). The fermented soydawadawa w ere low in tryptophan and tyrosine. B lackburn (1968) found that tryptophan and tyrosine are extensively destroyed on acid hydrolysis in the presence o f carbohydrates. H e suggested that appreciable amounts o f carbohydrates associated w ith the protein produce acid degradation products o f the sugars, such as hydroxymethyl furfural and that these may interact w ith amino acids to cause loss. Since soybean contains quite a high amount o f carbohydrates, about 40g/100g, it is expected that the loss o f amino acids during acid hydrolysis will be high. However, 77 University of Ghana http://ugspace.ug.edu.gh with fermentation o f the soybeans into soydawadawa, most o f the carbohydrates and available reducing sugars are utilized by microorganisms, reducing the loss o f amino acids. There were generally slight increases in the concentration o f the amino acids. This could be due to amino acids from the microorganisms as they increased during fermentation. With the use o f acid hydrolysis in this present work for the determ ination o f the protein profile, the trends in the above mentioned amino acids was well expected. M ethionine content was quite low in both unfermented soybeans and fermented soydawadawa. This is because methionine has been identified as the limiting amino acid in soybeans meals (Almquist and Grau, 1944). Eka, (1980) reported that dawadawa produced from African locust bean is low in the essential amino acids, leucine, isoleucine, phenylalanine and tryptophan. However, this work showed that dawadawa produced from soybean was low in only tryptophan with the other essential amino acids mentioned above comparatively higher The deficiency o f some o f the essential amino acids in dawadawa produced from African locust bean affects its value as a source o f high quality protein. Thus with soydawadawa improving upon this quality attribute, it could be regarded as a good source o f high quality protein in the meal. M oreover, the amino acids o f the main meal will not be solely relied upon to complement the low levels o f essential amino acids in traditional dawadawa. The amino acid content o f the Roasted Soydawadawa was slightly lower than that o f the Boiled Soydawadawa. Normally, dry heating is less effective than steam in enhancing the nutritional value o f soybean protein, and the presence o f w ater partially prevents the adverse effect o f excessive heating (Harris and Von Loesecke, 1960). 78 University of Ghana http://ugspace.ug.edu.gh This phenomenon may account for the observation that the Roasted Soydawadawa had a slightly lower amino acid content. There was a decrease in aspartic and glutam ic acids up to 48 h o f fermentation. Fetuga el. al. (1973) also observed similar results in the fermentation o f African locust bean dawadawa. 79 University of Ghana http://ugspace.ug.edu.gh Table 4-10: Am ino Acid Profile o f fermented and unfermented Soybean Dawadawa (mg/aa) FERM ENTED AM INO BOILED SOYDAW ADAW A ACID Oh 24h 48h 72h F.P ROASTED SOYDAW ADA Oh 24h 48h 72h F.P* UNFERMENTED SOYBEAN Isoleucine 228 227 232 280 276 225 223 230 272 269 230 Leucine 426 422 434 490 487 422 421 432 466 464 430 Lysine 338 336 342 400 398 332 331 341 392 390 340 Methionine 76 72 78 80 78 72 67 62 77 76 80 Cysteine 67 65 64 66 65 64 63 67 96 94 70 Phenylalanine 272 267 274 310 310 263 257 270 303 303 270 Tyrosine 207 205 213 200 198 203 200 192 196 193 210 Threonine 232 228 248 240 240 210 198 242 230 230 240 Tryptophan 89 86 89 80 78 85 83 79 77 75 90 Valine 302 300 312 300 300 280 276 335 289 288 310 Arginine 262 261 274 350 335 250 247 273 335 333 270 Histidine 147 144 153 160 157 132 120 154 158 157 150 Alanine 278 269 264 270 270 268 262 282 268 268 280 Aspartic 335 322 325 330 328 325 312 310 321 318 750 Glutamic 257 243 258 270 265 263 242 234 272 267 670 Glycine 233 229 246 260 256 238 235 242 252 250 240 Proline 229 218 338 340 335 229 226 234 339 336 240 Serine 324 318 316 320 320 298 296 312 318 316 330 F .P . - final p ro du c t 80 University of Ghana http://ugspace.ug.edu.gh 4.12 CHARACTERISATION OF PROTEINS IN SOYDAW ADAW A Hydrolysis o f protein in soydawadawa during fermentation was investigated by electrophoretically determining the protein profiles o f samples. The protein profiles are shown in Figure 4-8 and the profiles could be used to distinguish between the different fragments o r types o f protein-Lanes 3-11 showed bands which were less clear than those o f lanes 13-25. The experimental procedure in which soybeans were defatted before fermenting resulted in more pronounced breakdown o f the proteins than the treatm ent in which soybeans w ere fermented before defattening. The bands observed could have possibly resulted from the breakdown o f soy proteins into soy amino acids by proteinase enzymes. This occurs during fermentation o f soydawadawa since the key mechanism which occurs is the production o f proteinase by the dominant Bacillus species. As stated earlier, the number o f dominant m icroorganisms present in the fermenting soydawadawa increased with the progress o f fermentation. Their cumulative enzymatic activities also increased simultaneously, especially their proteolytic activity. Thus more enzymes or proteinases were produced to breakdown the protein in the soydawadawa, making available more amino acids. This supports the results obtained by the electrophoresis. The protein profiles o f both types o f soydawadawa during fermentation showed an increase in the fraction o f lower molecular weight proteins confirming the hydrolysis o f proteins during fermentation The protein bands (fig 4-7) became more distinct with the progress o f fermentation. This could be attributed to specificity in attack o f proteases on the proteins, resulting in the accumulation o f protein fragments o f certain lengths. 81 University of Ghana http://ugspace.ug.edu.gh Fig. 4-8 Protein profile o f soydawadawa at d ifferent stages o f ferm entation 1= High molecular weight protein (standard). 2= Low molecular weight protein (standard) 3= Raw soybean 4= Boiled but not fermented 5, 6, 7= Boiled and fermented for 24h, 48h, 72h. 8= Roasted but not fermented 9, 10, 11= Roasted and fermented for 24h, 48h, 72h. 12= Roasted but not fermented 13= Raw soybean defatted and fermented 14, 15, 16= Roasted and 17= Boiled but not fermented 18,19, 20= 21= Boiled but not fermented 22= Boiled, defatted, boiled 23; 24 ,25= Boiled and fermented fo r 24h, 48h, 72h Lanes 4 - 1 2 Beans w ere fermented before defattening Lanes 1 3 - 2 6 Beans were defatted before fermenting. University of Ghana http://ugspace.ug.edu.gh University of Ghana http://ugspace.ug.edu.gh 4.13 AROM A PROFILE OF SOYDAW ADAW A Table (4-11) shows that a lot o f difference were observed in the numbers and types o f aroma compounds present in the two different types o f soydawadawa suggesting that the aroma o f soydawadawa is greatly affected by the method o f production i.e. heat pretreatment. M ore aroma compounds w ere detected in the roasted soydawadawa and the compounds were also present at quite high levels. A few compounds however were detected in both the roasted and boiled soydawadawa. Acetophenone and pyrazines were found to be components related to protein degradation during maillard browning in roasted foods. This may explain the presence o f pyrazines in the roasted soydawadawa. Octadecanoic and hexanol were detected in both samples o f soydawadawa throughout the fermentation. This could have resulted from the hexane which was used as the solvent to extract the aroma. Tetradecanoic acid was detected in only the roasted sample throughout the period o f fermentation. On the whole, aroma compounds detected were basically, alcohols, ketones, aromatic and aliphatic organic acids and phenols. Table 4-12 also shows that some aroma compounds were produced as a result o f the fermentation. In the boiled soydawadawa, phenyl ethyl alcohol and 2,4-bis (1, 1 dimethyl ethyl phenol) which were not detectable at the start o f fermentation were detected in the fermented product. Similarly, in the roasted soydawadawa 1, 2- benzenedicarboxylic acid, 2-methoxy-phenol and 2, 5-dimethyl pyrazine were produced during fermentation because even though they were present in the fermented product they were not detected at the start o f fermentation. 83 University of Ghana http://ugspace.ug.edu.gh The flavour impressions o f many foodstuffs implies a complex m ixture o f aroma substances and that many traditional foodstuffs and beverages are flavoured in situ by the action o f m icroorganisms e.g. the use o f proteases produced by bacteria in the meat industry (Schermers el al. 1976). Many vegetable proteins such as soybeans possess characteristic beany flavours and odour which limit their sale and use because the flavour and/or odour is unappealing to large numbers o f potential consumers. Non pathogenic bacteria have however been found to proliferate such foods and remove the flavour and/or odour from it (Hanson, 1974). W ang et at. (1968) found that bacteria o r moulds during the fermentation o f soybeans into various foodstuffs help to improve the flavour and perhaps also add to their nutritional value. Honig et. al. (1976) found roasting to be necessary in order to produce soy protein p roducts with flavour scores approaching the blandness o f wheat flour. Although flavour scores w ere not done on the soydawadawa, one could perhaps deduce from the aroma compounds produced that the roasted product was more flavourful than the boiled one. Hesseltine, (1965) also found fermentation to improve or modify flavour, taste and texture o f foods. Vandore, (1967) found dry method o f cooking such as grilling and roasting to produce meats o f superior flavour. 84 University of Ghana http://ugspace.ug.edu.gh Table 4-11: M ajor aroma compounds detected by GC -M S during the fermentation o f soydawadawa. BOILED ROASTED A R O M A C O M PO N EN T SOYDAWADAWA SOYDAWADAWA Oh 24h 48h Oh 24h 48h 3 - Hexanol + + + + + + 9, 12 - O ctadecanoic acid + + + + + + Phenyl ethyl alcohol - + + - 2, 4-bis (1 ,1 -dimethyl ethyl phenol) + + + Tetradecanoic acid - + + + 1, 2-Benzenedicarboxylic acid - - + + 2 - methoxy - phenol - - + 2, 5 - dimethyl pyrazine - - + 85 University of Ghana http://ugspace.ug.edu.gh 5.0 CONCLUSIONS The heat pretreatm ent given to soybeans before processing may favour the selection o f heat resistant Bacillus spores and hence makes them the predominant m icro-organisms responsible for the fermentation o f soybeans into dawadawa. The Bacillus species identified in this work are mainly, Bacillus subtilis, Ii licheniformis, B. pumihis, B .firmus and B, cereiis. This confirms the work o f previous workers who had also noted the dom inance o f Bacillus species over o ther m icro­ organisms during dawadawa fermentation. Hydrolysis o f proteins during fermentation has been confirmed by an increase in the fraction o f lower molecular weight proteins in the protein profiles o f both types o f soydawadawa. The attack o f proteases on the proteins also increased with the progress o f fermentation. This is evidenced by more distinct protein bands in the electrophoresis gels as fermentation progressed. This phenomenon is due to the accumulation o f protein fragments o f certain lengths. The Bacillus species that were identified showed proteolytic activity and are important for the proteolysis associated with soydawadawa production. Some products o f this proteolysis are responsible for the characteristic strong smell o f soydawadawa. The pH o f the fermenting beans confirms the alkaline nature o f the process. The compounds responsible for the alkalinity could be mainly volatile nitrogenous compounds because the dried final products do not have the high pH found in the fermenting mass. This alkalinity enhances the grow th o f the Bacillus species. The form o f heat used for the pretreatment had a significant elTect on the aroma compounds and hence the aroma o f the resulting end product. This work identified the 86 University of Ghana http://ugspace.ug.edu.gh main organic acids and volatile aroma compounds developed during the fermentation process. The major aroma compounds identified included 9 ,12-octadecanoic acid, phenyl ethyl alcohol, 2, 4 -bis (1, 1-dimethyl ethyl phenol), Tetradecanoic acid, 1 , 2 - Benzene dicarboxylic acid, 2 - methoxy - phynol and 2, 5 - dimethyl pyrazine 87 University of Ghana http://ugspace.ug.edu.gh 6.0 RECOMMENDATIONS The identification o f Bacillus species as the dominant micro - organism s responsible for soydawadawa production and their enzymatic activities could serve as a first step in the process o f developing a starter culture. This starter culture could be used for producing soydawadawa with consistent high quality. It is also recommended that technological properties such as proteinase and the a - amylase activities o f the starter culture should be typed. Optim ization o f the fermentation param etres should be done to reduce fermentation time for the production o f soydawadawa. 88 University of Ghana http://ugspace.ug.edu.gh REFERENCES A bbiwD . 1990. Useful plants o f Ghana: Intermediate Technology Publications Ltd; London, PP 48-254. Adewuyi. E.Y. 1983. Studies on the Optim ization o f process conditions for locust bean (.Parkia biglobosa) fermentation, MSc dissertation. Department o f Botany, University o f Ibadan. Almquist, H. J. and Grau, C. R. 1944. Amino Acids in Soybean Varieties. Poultry Science. 23, 34. Alozie, T. C., Rotimi, C. N. and Oyibo, B. B. 1980. Production o f aflatoxin by Aspergillus flavus (UBMI) in some N igerian indigenous beverages and foodstuffs. M ycopathol. Mycol. Appl. 70, 125. Annegers, J. F. 1974. Protein quality o f West African foods. Ecol. Food Nutr. 3, 125-130. Antai, S. P. and Ibrahim, M. H. 1986. M icroorganisms associated with African locust bean (Parkia filicoidea welw.) fermentation for dawadawa production. J. Appl. Bacteriol. 61, 145-148, AOAC (1990) Official M ethods o f Analysis, 15ih edn. Virginia. Association o f Official Analytical Chemists. Arcega, B.1969. The M anufacture o f Sauce, N IST Tech. Bulletin. N o.9. Barimalaa, I.S., Achinewhu, S.C., Yibatama, 1. & Amadi, E.N. 1994. Studies on the solid substrate fermentation o f soybeans (Glycine max var Merril). J. Appl. Bacteriol. 80. 89 University of Ghana http://ugspace.ug.edu.gh Bemelmans, J. M. H. 1981. Isolation and concentration of volatiles from foods. Chpt.2 in “ Isolation, separation and Identification o f volatile compounds in Aroma research” , ed. H. M aarse and R. Belz, pp. 4-59. Akademie-Verlag, Berlin. Bernfeld, P 1955. Amylases and Proteinases. In M ethods in enzymology, vol 1. (Colow icz S. P. and Kaplan N. 0 . eds). Pp. 149-158. New York: Academic press. Blackburn, S. 1968. Amino Acid Determination, M ethods and Techniques. Marcel Dekker, Inc; New York Pp. 125. Busson, F. 1965. Plantes Alimentaires de l’ouest African, D irection de la Cooperation Culturelle et Technique. Paris, p. 272. Campbell- Platt, G. 1987. Fermented foods o f the world, A Dictionary and Guide. London; Butterworths. Campbell-Platt, G. 1980. African locust bean (Parkia species) and its West Africa fermented food product, dawadawa. Ecology o f Food and Nutrition. 1:123 - 132. Carrao, M. C., Panizzi, J. M ., Gontijo M andarino, G. 1994. Soybean for Human consumption, Nutritional quality, processing and utilization. FAO, Rome pp. 241 - 253. Claus, D. and Berkerly, R. C. W (1986). The genus Bacillus. In Bergey’s Manual o f Systematic Bateriology, Vol. 2. Ed. Sneath, P.H.A, Mair, N. S, Sharpe, M. E. and Holts, J. G. pp. 1 105-1139, Baltimore: Williams and Wilkins. Dema I. S. 1965, Nutrition in Relation to Agricultural Production. Food and Agriculture Organization. Rome. 129. D iawara, B. Sywadogo, L. Amoa-Awua, W. K. A. and Jakobsen, M. 1998. Quality system for the production o f soumbala. The HACCP System. Taastrup: Waitro. 90 University of Ghana http://ugspace.ug.edu.gh Eka, 0 . U. 1980. Effect o f fermenta tion on the nutrient status o f locust beans. Food Chem istry. 5 : 303-308. FAO Food and Nutrition Paper 47/4 (1989). Utilization o f tropical foods: Tropical beans FAO, Rome. Fetuga B. L., Babatunde, G. M. and Oyenuga, V. A, 1973. Protein quality o f some Nigerian Foodstuffs. I. Chemical assay o f nutrients and amino acid composition. J. Sci. Food Agric. 24, 1505. Forgarthy, W. M. & Griffin, P. J. 1973. Production and Purification o f metalloproteases o f Bacillus polymyxa. Applied M icrobiology 26, 191 -195. Frazier, W. C. (1967) Food M icrobiology. 2nd ed. New York: Me G raw - Hill, p. 450. Girgit, P and Turner, T. D. 1972. Lesser known Nigerian edible oils and fats. III. Fatty acid composition as determined by gas-liquid Chromatography, J. Sci. Food Agric. 23-25. Gonzalez, O. N. and Uyenco, F R. 1975. M icrobiology o f Fermented Foods. Paper presented at the technical seminar. 2011' July - 2,ul August. Manila: ASEAN Sub-committee on Protein Gordon, M. H. and Macrae, R. 1992. Instrumental Analysis in the Biological Sciences. Blackie and Sons Ltd. Edmunds, Suffolk, p. 41-43. Gordon, R. E. Haynes, W. C. and Pang, C. H. N. (1973) The genus Bacillus, Handbook N“ 427. Washington, United States Department o f Agriculture. Gordon, R. E (1973) The genus Bacillus. In: Laskin A. I and . Lechevalier H A. (Eds.), CRC Handbook o f M icrobiology. Vol. 1, Organismic Microbiology. CRC Press, Cleveland. OH. Pp. 71-81. 91 University of Ghana http://ugspace.ug.edu.gh Hanson, L. P 1974. Vegetable Protein Processing., Pg. 25-26. Noyes data Corporation, New Jersey Harris, R. S. and Von Loesecke, H. 1960. Nutritional Evaluation o f food Processing. Pp. 239. John Wiley. Hesseltine, C. W. 1965. A millennium o f fungi, food and fermentation. Mycologia 57: 149-197. Hesseltine, C. W„ Smith, N ., Bradle, B. and Djien, L. 1963. Investigations o f Tempeh and Indonesian Food. W ashington D. C: American Institute o f Biological Sciences. Honig, D. H., Warner, K. and Rackis, J. J. 1976. Soybean Protein flavour. J. Food Sci. 4 1 ,642 -646 . Horii, M. 1997. Soybean and Fermented Food culture in the world. J. Farming Japan Vol. 31-4, pp. 10-20. Hug, R. S., Abalaka, J. A. and Stafford, W. L. 1983. Folate Content o f various Nigerian foods. J. Sci. Food Agric, 34, 404, 1983. Hugh, R. and Leifson, E. 1953. The taxonom ic significance o f fermentation versus oxidative metabolism o f Carbohydrates by various gram negative bacteria. J. Bacteriol. 66, 24-26. Ihekoranye, A. I and Ngoddy, P. 0 . 1992. Integrated Food Science and Technology for the Tropics; The Macmillan Press Ltd., London; pp, 92. Ikenebomeh, M. J. (1989). The influence o f salt and temperature on the natural fermentation o f African Locust bean. 92 University of Ghana http://ugspace.ug.edu.gh Ikenebomeh, M. J. 1982. The solid substrate Fermentation o f African Locust Beans {Parkia filicoidect) Ph. D. thesis, Me Gill University (McDonald Campus) Montreal. Ikenebomeh, M. J., Kok, R. and B lackwood, A. C. 1981. Solid substrate fermentation o f the African locust bean (Parkiafi/icoidea welw.), Abstract. SIM News. 31 -46. Jennings, W. G. 1980. Sample preparation. Chpt. 12 in “Gas chromatography with glass capillary columns,” 2nd edition., pp, 183 - 200. Academic Press, N ew York. Keshinro, O. 0 . 1983. The Free and Total folate activity in some commonly available foodstuffs, J. Food Chem. p. 11, 87. Lawson, R. M. 1965. The Consumption approach to measuring Agricultural production in foodstuffs. Food Res. Inst. Stud. J. Agric. Econ. T rade Dev. (Standford). p. 5, 205. Lee, S. Y. Park, K. W. and Min, Y. K. (1983) Nutritional evaluation o f naturally fermented soybean and the enzymatic activity changes during the preparation. Korean Journal-of-food-science-and-Technology. 15 (2) p. 101-107. Lilie, R. D. (1928) The Gram Stain. 1. A quick method for staining Gram positive organisms in the tissue. Archives o f Pathology 5, 828. Logan, N. A. and Berkerley, R. C. W. (1984) Identification o f Bacillus strains using API system. J. Gen. M icrobiol. 130 - 1 8 7 1 - 1 8 8 2 . Logan, N, A. and Berkerley, R. C. W. (1981) Classification and identification o f the genus Bacillus using API tests. In The Aerobic Endospore forming Bacteria: Classification and Identification. Pp. 105-140. London: Academic Press. Kordylas, M. 1991. Processing and Preservation o f Tropical and Sub-Tropical Foods. English Language Book Society, Hong Kong. pp. 52-58. 93 University of Ghana http://ugspace.ug.edu.gh Nicol, B. M. 1959. The protein requirements o f Nigerian peasant farmers. Br. J. Nutr. 13, p7. Nikkuni, S. 1997. Nutritive value and processing o f soybeans. Food Science & Technology. Int.; 90-93. Nikkuni, S., Karki, J.B. Vilkhu, K. S., Suzuki, T., Shindoh, K., Suzuki, C. and Okada, N. 1995. Food Science Technology, Int. 1, 107-111. Nnam, N. M. 1995. Evaluation o f nutritional quality o f fermented cowpea (vigiia imguiculata) flours. Ecology-of-food-and-Nutrition (USA). V 33 (4) p. 273-279, Tables, 32, ref. Obizoba, I. C. and A tu-L.N . 1993. Production and chemical evaluation o f some food condiments o f Nigeria. P lant-Foods-for-Human Nutrition. 44:3, 249-254; 2 ref. Odunfa, S. A. 1985. Biochemical changes in fermenting African locust bean (Parkin biglobosci) during iru fermentation. J. o f Food Technology. 20, 295-303. Odunfa, S. A. 1983. Carbohydrate changes in fermented locust bean (Parkia filicoidea) during iru preparation. Plant foods for Human Nutrition 32: 3- 10. Odunfa, S. A. 1981. M icro-organisms associated with fermentation o f African locust bean {Parkiafilicoidea) during iru preparation. J. Plant Foods 3: 245-250. Odunfa, S. A. and Adesomoju, A. A. 1983. Effects o f fermentation on the free fatty acids o f African locust bean during iru preparation. J. Plant foods. Odunfa, S. A. and Oyewole, B, O. 1986. Identification o f Bacillus Species from iru, a fermented African locust bean product. J. Basic Microbiol. 26, 101-108. Odusote, K. O. 1980. M icrobiology o f melon fermentation for “ogiri” production. R eport in Department o f Botany, University o f Ibadan, Nigeria. Pp. 28. 94 University of Ghana http://ugspace.ug.edu.gh Ogbadu, L. J. and Okagbue, R. N. 1988. Fermentation o f African locust bean (.Parkia biglobosa) seeds: involvement o f different species o f Bacillus. Food M icrobiol. 5, 195-199. Ogunbunmi, E. M. and Bassir, 0 . 1980. Protein and amino acid contents o f some N igerian food Condiments. Nutr. Rep. Int. 22, 497. Oke, O. L. 1969. Oxalic acids in plants used in nutrition. World Rev. Nutr. Diet. I 1. 170. Oke, 0 . L. 1972. A nutrition policy o f Nigeria, World Rev. Nutr. Diet. 14, I . Oxoid, 1979. The Oxiod Manual, 4Ul edn. Oxiod Ltd. Hampshire, U. K., 310 pp. Paredes-Lopez, 0 and Harry, G. I. 1988. Food biotechnology review: traditional solid-state fermentation o f plant raw materials. Critical-reviews-in food-science- and-nutrition. (USA). V. 27 (3) p -159-187. Pariy, J. M. Turnball, P. C. B. and Gibson J. R. 1983. A colour Atlas o f Bacillus Species. Ipswich: Wolfe Medical Pub. Ltd. Pederson, C. 1971. M icrobiology o f Fermented Foods. W estport: AVI, Potter N. N. 1968. Food Science: The Avi Publishing Company, Inc. West Port, Connecticut. Priest, F and Austin, B. 1993. Modern Bacteria Taxonomy, second edition. London: Chapman Hill. Rose, A. H. 1982. Fermented Foods. Econom ic M icrobiology Volume 7. Great Britain Alden Press, Oxford P. 30, 45-80. Sakar, P. K., Tamang, J. P ., Cook, P. E. and Owens, J. D. 1994. Kinema - a traditional soybean fermented food: proximate composition and microflora. Food M icrobiol. 11, 47-55. 95 University of Ghana http://ugspace.ug.edu.gh Sakar, P. K. and Tamang, J. P. 1994. The influence o f process variables and inoculum composition on the sensory quality o f kinema. Food m icrobiology 11 1, 3 17 - 325 Schermers, F. H., DufEus, J. H. and Macleod, A. M. 1976. The Role o f M icroorganisms in F lavour Formation. J. Inst. Brewing, 82, 170. Schery, R. W. 1972. O ther food seeds and forages. In: Plants for Man. 2n