Characterization of Killer Yeasts from Nigerian Traditional Fermented Alcoholic Beverages

Adesokan I. A. The Polytechnic Journal of Science and Technology PJST. Vol. 8. 2013 ISSN: 1115 8336 2013 Journal of Science & Technology Polytechnic Ibadan, Nigeria Characterization of Killer Yeasts from Nigerian Traditional Fermented Alcoholic Beverages Adesokan I. A. Department of Science Laboratory Technology, The Polytechnic, Ibadan, Nigeria. Email address: sokanisaac@yahoo.co.uk Abstract In this study the killer phenotypes among yeasts isolated from Nigerian traditional alcoholic beverages was investigated using methylene blue medium. The ethanol tolerance of the yeast isolates was also determined. A total of 62 yeasts were identified and found to belong to 16 different species and 10 genera. Saccharomyces cerevisiae was isolated from all the beverages and had a percentage occurrence of 33.33% in burukutu. It was closely followed by members of the genus Candida with 26.67% occurrence in the same beverage. Other yeasts isolated were S. chevalieri and S. uvarum. The viable counts of the yeasts ranged between 2.0x107 and 9.5x10 5 cfu/ml while the ph was between 3.6 and 5.4. Candida tropicalis and C. castelli were the most ethanol tolerant by having a growth of 1.5% in broth with 10% ethanol. All the strains of S. cerevisiae tested exhibited killer activity. Other killer yeasts isolated were members of genus Candida (except C. fructus), Pichia ohmeri, Schizosaccharomyces japonicum, S. chevalieri, Geotrichum candidum and Kluveyromyces apiculata. The non-killer yeasts isolated were S. uvarum, Sch. pombe, Rhodolorula graminis, K. apiculata and Pichia membranefaciens. P. membranefaciens and R. graminis were neutral while others were sensitive. Although all the strains of S. cerevisiae tested were killer yeasts, strain PW1 was especially ethanol tolerant and therefore seems to be a potentially useful strain. Introduction Fermented foods and beverages constitute a very important component of the people s diet in Africa. There are several different fermented products, which have been reported. Many fermented foods serve as main course meals and such food include gari, fufu and ogi. Others are highly prized food condiments. Those that are food condiments are usually made from the fermentation of protein rich seeds. Some of these are iru from Africa locust bean; Ugba from Africa oil bean and ogiri from melon seeds (Odunfa and Oyewole, 1998). There are five different types of traditional alcoholic beverages that are well known in Nigeria. These are burukutu, pito, sekete, which are prepared from malted maize sorghum grain, PJST Vol. 8, 2013 13

Characterization of Killer Yeasts from Nigerian Traditional Fermented Alcoholic Beverages agadagidi from ripe plantain pulp and palm-wine from sugary sap of palm tree (Elaeis guineensis) and raphia palm (Raphia hookeri) (Sanni and Lonner, 1993). These beverages apart from serving as inebriating drinks are also important in fulfilling social obligations such as marriage, naming and burial ceremonies. Their production is as a result of mixed, spontaneously generated fermentation with yeasts and lactic acid bacteria as the predominating microorganisms. However, the yeasts flora is mainly responsible for the alcohol content of the beverages (Sefa-Dedeh et al., 1999). The production by yeasts of exotoxins with antimicrobial activity mediated by specific cell wall receptor on susceptible microorganisms is a relatively common phenomenon (Polonelli and Morace, 1986; Polonelli et al., 1991). Exotoxins (generally protein or glycoprotein) that are able to kill susceptible cells belonging to the same or congeneric species have been defined as killer toxins. Killer yeasts are toxin-producing fungi that are immune to the activity of their own killer toxins. After the original description of the phenomenon in Saccharomyces cerevisiae reported by Bevan and Makower (Yap et al., 2000), the attention of increasing number of investigators has been focused in the killer effect, which might represent a model of biological competition somewhat related to that of bacteriocin among bacteria (Musmanno et al., 1999; Schmitt and Breinig, 2006; Buzzini et al., 2007). Various yeasts species have been reported in traditional fermented alcoholic beverages in Nigeria (Sanni and Lonner, 1993) while the killer activities of these different yeasts species have been scarcely reported. Therefore, the objective of this present study is to provide information on the killer, sensitive and neutral phenotypes among the yeast isolated from Nigerian traditional fermented alcoholic beverages. Materials and Methods Sample collection Palm wine, burukutu and pito used in this study were from the local producers within Ibadan metropolis and agadagidi was from Ile-ife. Four samples of each beverage were collected separately in sterile containers and kept at 4 o C before analysis. Sensitive strain Candida glabrata Y55 and killer strains P. kluyveri UW2 and Kluyveromyces UW1 were obtained from the culture collection of Department of Biology, University of Western Ontario, Canada. Isolation and identification of yeasts Isolation was carried out on samples of each of these alcoholic beverage collected at various interval. One ml of suitable dilutions of each sample was poured into sterile Petri dishes and sterile molten (45 o C) Potato Dextrose Agar (PDA) containing 30µg/ml streptomycin was poured into the Petri dishes. The plates were purified by repeated streaking on PDA. The purified cultures were streaked onto PDA slants, incubated at 25 o C for 48 hours. The yeast isolates were then identified by physiological and biochemical tests according to the scheme described by Kreger-Van Rij (1987). PJST Vol. 8, 2013 14

Adesokan I. A. Viable counts and ph measurement The viable count was determined using the pour plate technique. Suitable dilutions of each sample were plated in duplicates and colony forming units was obtained on PDA after incubation at 25 o C for 72 hours. The ph of the samples was determined using a ph meter. Determination of killer activity Killer phenotype was determined according to the method of Guitierrez et al. (2001). The assay medium was yeast extract peptone dextrose (YEPD) agar containing 0.03g/l methylene blue and buffered to ph 4.5 and 3.5 with 0.2M citrate-phosphate buffer. All isolates were transferred to plates seeded with sensitive strain Candida glabrata Y55 and incubated at 25 o C for 48-96 hours. The tested strain was designated a killer yeast if the streaked strain was surrounded by a region of bluish-stained cells or by a clear zone of growth inhibition bounded by stained cells. To test the sensitive and neutral phenotypes, killer strains Kluyveromyces sp UW1 and Pichia kluyveri UW2 were streaked onto plates seeded with each of the non-killer isolates. If the killer strain was surrounded by a clear zone of inhibition or by a region of bluish-stained cells, then the tested strain was designated as sensitive. If there was no clear zone of inhibition or a region of bluish-stained cells then the tested strain was designated as neutral. That is a sensitive strain was inhibited by the killer strain while a neutral strain was not. Ethanol tolerance Ethanol tolerance was measured as the ability of the yeast to grow in yeasts extract peptone dextrose (YEPD) broth with increasing content of ethanol 5%, 10% and 15%. After 72 hours of incubation, growth was measured by turbidity increase at 650nm with camspec M105 spectrophotometer (Sefa-Dedeh et al., 1999). Results From the four different samples of each alcoholic beverages analyzed a total of 62 yeast isolates were identified. The yeasts were found to belong to 16 different species and 10 different genera (table 1). The most frequently encountered yeast which was found in all beverages was Saccharomyces cerevisiae. Also S. chevalieri was found in all the beverages except in palm wine while S. uvarum was isolated from burukutu and pito. The members of the Genus Candida isolated were Candida valida, C. krusei, C. intermedia C. tropicalis and C. castelli. Each of these yeasts was isolated from one beverage except C. valida which was found in both burukutu and pito. The viable counts of the yeasts from the beverages ranged between 2.0X 10 7 to 9.5X10 8 cfu/ml while the ph was between 3.6 and 5.4 (table 2). The ethanol tolerance of the yeasts (figure 1) showed that Candida tropicalis and P. membranefaciens had the highest growth of about 1.5% in YEPD broth with 10% ethanol. Whereas S. cerevisiae BK2, PT1,PW1, G. candidum, Sch. japonicum, Sch. pombe and R. graminis had a growth above 1.0% in YEPD broth with 10% ethanol. Furthermore, Candida valida, S. cerevisiae BK2, PW1, S. chevalieri, R. graminis, C. castelli and C. fructus grew in all the 3 ethanol concentration used. PJST Vol. 8, 2013 15

Characterization of Killer Yeasts from Nigerian Traditional Fermented Alcoholic Beverages However, P. kluyveri UW2 and Kluyveromyces UW1 grew only in YEPD broth with 5% ethanol. Meanwhile, all the yeasts that grew in YEPD broth with 15% ethanol had a value below 0.5% except C. intermedia which was above 0.5%. The killer activity of the yeast isolates are presented in table 3. All the strains of S. cerevisiae tested and the members of genus Candida isolated (except C. PJST Vol. 8, 2013 16 fructus) exhibited killer activity. Other killer yeasts isolated were P. ohmeri, Sch. japonicum, S. chevalieri, G. candidum and K. africanus. The non-killer yeasts isolated were S. uvarum, Sch. Pombe, R. graminis, K. apiculata and P. membranefaciens. P. membranefaciens and R. graminis were neutral while others were sensitive. Table 1: Distribution of yeast species found in the alcoholic beverages Species Bukurukutu Pito Agadagidi Palm wine No % No % No % No % Candida valida 4 26.67 1 6.67 - - - - Saccharomyces cerevisae 5 33.33 3 20.00 5 26.32 3 23.08 Saccharomyces 1 6.67 3 20.00 1 5.26 - - chevalieri Pichia ohmeri 2 13.33 3 20.00 - - 2 15.38 Geotrichium 1 6.67 - - - - - - candidum Saccharomyces 1 6.67 1 6.67 - - - - uvarum Schizosaccharamyces 1 6.67 1 6.67 - - 1 7.69 japonicum Kluyveromyces - - 2 13.33 2 10.53 - - africanus Schizosaccharomyces - - - - 1 5.26 - - pombe Candida krusei - - - - 4 21.05 - - Candida intermedia - - - - 1 5.26 - - Kloeckera apiculata - - - - 1 5.26 1 7.69 Rhodolorulagraminis - - - - 3 15.78 2 15.38 Candida tropicalis - - - - - - 1 7.69 Candida castelli - - - - - - 1 7.69 Pichia - - - - - - 1 7.69 membranefaciens Torulaspora - - - - - - 1 7.69 delbruekii - - 1 6.67 1 5.26 - -

Adesokan I. A. Table 2: Yeasts viable counts and ph of the beverages Beverages ph Colony forming unit/ml Agadagidi 3.6* 2.0 x10 7 Burukutu 3.6 5.3x10 7 Palm wine 5.4 9.5x10 8 Pito 3.8 7.2x10 8 *Values are means of four determinations. Discussion In this study yeasts were isolated from four different Nigerian traditional fermented alcoholic beverages. All the isolated yeasts in this study have been reported with exception of C. valida. Sefa-Dedeh et al. (1999) characterize yeasts in traditional brewing of pito in Ghana and identified 21 strains belonging to 8 genera as Saccharomyces cerevisiae (8), Candida tropicalis (4), Kloeckera apiculata (2), Hansenula anomata (2) among others. Theivendirarajah and Chrystopher (1986) reported three general of yeasts in palmyrah palm wine. They stated the organisms of significant are those belonging to the genus Saccharomyces, because of their numerical predominance and superior fermentative ability. In the present study members of the genus Saccharomyces had the highest percentage of occurrence and were isolated from all the alcoholic beverages studied. This showed that the member of this genus is the most predominant and the result is in agreement with the observation of earlier works (Ekundayo, 1969; Theivendirarajah and Chrystopher 1986; Demuyakor and Ohta, 1991; Sanni and Lonner, 1993). PJST Vol. 8, 2013 17

Characterization of Killer Yeasts from Nigerian Traditional Fermented Alcoholic Beverages Table 3: Killer activity of yeasts isolated from Nigerian traditional fermented alcoholic beverages Species Sensitive yeast in the lawn Killer activity (indicator organism) Pichia ohmeri Candida glabrata + Candida fructus Candida glabrata - C. krusei Candida glabrata + C. valida Candida glabrata + C. tropicalis Candida glabrata + C.intermedia Candida glabrata + Schizosaccharomyces pombe Candida glabrata - Sch. Japonicum Candida glabrata + Saccharomyces cerevisiae Candida glabrata + AG19 S. cerevisiae BK2 Candida glabrata + S. cerevisiae PT1 Candida glabrata + S. cerevisiae PW1 Candida glabrata + S. uvarum Candida glabrata - S. chevalieri Candida glabrata + Geotrichum candidum Candida glabrata + Kluyveromyces africanus Candida glabrata + Kloeckera apiculata Candida glabrata - Rhodolorula graminis Candida glabrata - Pichia membranefaciens Candida glabrata - Torulaspora delbruekii Candida glabrata - PJST Vol. 8, 2013 18

PJST Vol. 8, 2013 19 Adesokan I. A.

Characterization of Killer Yeasts from Nigerian Traditional Fermented Alcoholic Beverages Next in terms of numerical number are the members of the genus Candida. The Candida isolated are C. valida, C. krusei, C. intermidia, C. tropicalis, C. castelli and C. fructus. All these members have been reported by previous studies (Sanni and Oso 1988). The only exception is Candida valida. The yeast viable counts and ph range reported in this work was significantly different from values reported by Sanni and Lonner (1993). A ph range of 3.4 to 5.4, and yeast viable counts of 8.4x10 4 to 6.2x10 6 cfu/ml was reported by these workers. The unpredictability of the associated yeast flora of traditional alcoholic beverages in sub-sahara Africa could be due to spontaneous nature of the fermentation, sources and type of ingredients used. The growth of the test isolates at different ethanol concentration over a period of time indicates their capacity for ethanol tolerance (Sefah-Dedeh et al., 1999) even though the alcoholic content of traditionally brewed beverages is between 1.5-6% (Odunfa and Oyewole, 1998). Although ethanol tolerance is not directly related to ethanol production capability (Sanni and Onilude, 1998), such information could be important for selection of potential strains of yeasts that can be employed for optimization of processing of the traditional alcoholic beverages with higher ethanol content. The dependence of killer activity on ph was first observed by Woods and Bevan (1986). Since then, the optimum ph for killer activity has been defined for a number of killer toxins, with the majority of them having an optimum ph between 4.2 and 4.7 (Pfeiffer and Radler, 1984). It was observed that all the strains of S. cerevisiae tested exhibited killer activity at ph 3.5 and 4.5. However, the killer activity of C. castelli, C. tropicalis and G. candidum was not detectable at ph 3.5 (result not shown). All the strains of S. cerevisiae isolated were killer yeasts and they showed a very strong ethanol tolerance. S. cerevisiae PW1 and BK2 grew at 15% ethanol concentration whereas PT1 and AG19 did not grow at all. Sefa-Dedeh et al. (1999) reported that S. cerevisiae I was the most ethanol tolerant while K. africanus was the least. Several studies have shown different applications of killer yeasts for biotyping, wine fermentation, potential antifungal activity (Perez et al., 2001; Sangorin et al., 2002; Schmitt and Breinig, 2002; Buzzini et al., 2007). Although all the strains of S. cerevisiae tested were killer yeasts, strain PW1 was especially ethanol tolerant and therefore seems to be a potentially useful strain. Further investigation could be carry out to determine the strong and weak killer phenotypes among the yeasts so as to pick the one with best toxin production ability and whose toxin has a broad spectrum of activities against many yeasts. PJST Vol. 8, 2013 20

ACKNOWLEDGEMENT The author gratefully acknowledge the technical support and positive criticism by Prof. A.I. Sanni of Department of Microbiology, University of Ibadan, Nigeria in the course of supervising this project and Prof. M.A Lachance of Department of Biology, University of Western Ontario, Canada for kind donation of some yeast isolates used in this research work. The technical support of Dr S.T. Ogunbanwo of Department of Microbiology, University of Ibadan, Nigeria was also appreciated. REFERENCES Buzzini P., Turchetti B and Vaughan-Martini A.E (2007). The use of killer sensitive patterns for biotyping yeast strains: the state of art, potentialities and limitation. FEMS Yeast Research 7: 749-760. Demuyakor B and Ohta Y (1991).Characteristics of single mixed culture fermentation of pito beer. Journal of Science of Food and Agriculture 62: 401-408. Ekundayo J.A (1969). The Production of Pito, a Nigerian fermented beverage. Journal of Food Technology 4: 217-225. Gutierrez A.R. Epifanio S., Garijo P., Lopez R. and Santamaria P. (2001). Killer yeasts: Incidence in the Ecology of Spontaneous fermentation. American Journal of Viticulture 52: 352-356. Kreger-Van Rij N.J.W (1987). The Yeasts a Taxonomic Study 3 rd Edition, Amsterdam, Elsevier Publishers. Adesokan I. A. Musmanno R.A., Di Maggio T. and Coratza G. (1999). Studies on strong and weak killer phenotypes of wine yeasts: production, activity of toxin in must, and its effect in mixed culture fermentation. Journal of applied Microbiology 87: 932-938. Odunfa S. A and Oyewole O.B (1998). African fermented foods in Microbiology of fermented foods. Vol. 2 2 nd ed. (B.J.B Wood, ed) Elsevier/Applied Science Publishers, London. Pp713-746. Pfeiffer P and Radler F. (1984). Comparison of the killer toxin of several yeasts and the purification of a toxin of type K2. Achives of Microbiology 137: 357-361. Polonelli L and Morace G (1986). Reevaluation of the yeast killer phenomenon. Journal of Clinical Microbiology 24: 866-869. Polonelli L., Lorenzini R., Bernardis F and Morace G (1991). Potential therapeutic effect of yeast killer toxin. Mycopathologia 96: 103-107. Sanni A.I and Lonner C (1993). Identification of yeasts isolated from Nigerian traditional alcoholic beverages. Journal of Food Microbiology 10: 517-523. Sanni A.I and Onilude A.A (1998). The effect of temperature on the ethanol tolerance of yeasts isolated from Nigerian traditional alcoholic beverages. Science Research Journal University of Ibadan, Nigeria 3:10-15. PJST Vol. 8, 2013 21

Characterization of Killer Yeasts from Nigerian Traditional Fermented Alcoholic Beverages Sanni A.I. and Oso B.A. (1988). The production of agadagidi, a fermented beverage. Die Nahrung 32: 319-326. Sefa-Dedeh S, Sanni A.I, Tetteh G and Sakyi- Dawson E (1999). Yeasts in the traditional brewing of pito in Ghana. World Journal of Microbiology and Biotechnology 15:015-020. Schmitt M. J and Breinig F. (2002). Yeast viral killer toxins: lethality and selfprotection. Reviews 4: 212-221. Schmitt M. J and Breinig F. (2006). The viral killer system in yeast: from molecular biology to application. FEMS Microbiology Reviews 26: 257-276. Sangorrin M., Zajonskovsky I., Broock M. and Caballero A (2002). The use of killer biotyping in an ecological survey of yeast in an old Patagonian winery. World Journal of Microbiology and Biotechnology 18: 115-120. Theivendirarajah K and Chrystopher R.K. (1980). Microflora and Microbial activity in palmyrah (Borassus flabellifer) palm wine in Sri Lanka MIRCEM Journal 3: 23-31. Woods D.R and Bevan E.A. (1986). The studies on the nature of the killer factor produced by Saccharomyces cerevisiae. Journal of General Microbiology 51: 115-126. Yap N. A., Lopes M. B., Langridge P. and Henschke P.A. (2000). The incidence of killer activity of non- Saccharomyces yeasts towards indigenous yeasts species of grape must: potential application in wine fermentation. Journal of Applied Microbiology 89: 381-389. Perez F., Ramirez M and Regodon J.A. (2001).Influence of killer strains of Scharomyces cerevisiae on wine fermentation. Antonie van Leeuwenhoek 79: 393-399. PJST Vol. 8, 2013 22