Relationship Between Fatty Acid Concentrations in Wine Yeasts and Sugar Fermentation at Different Temperatures *

Transcription

1 Relationship Between Fatty Acid Concentrations in Wine Yeasts and Sgar Fermentation at Different res * J.J. Venter!, R.J.J. van Vren 2, A. Tromp! and J.R. RandalP 'Viticltral and Oenologicai Research Institte (VORl), Private Bag X526, 76 Stellenbosch, 2Department of Microbiology and 3Department of Biometry, University of Stellenbosch, 76 Stellenbosch, Repblic of Soth Africa. Sbmitted for pblication: April 1989 Accepted for pblication: Agst 1989 Key words: Fatty acid, wine yeast, fermentation, temperatre The effects of temperatre and fermentation stage on the celllar concentrations of nsatrated and satrated fatty acids on the fermentation abilities offor different Saccharomyces cerevisiae wine yeasts were stdied at three temperatres. The total nsatrated and satrated fatty acid concentrations were determined at six fermentation stages. Unsatrated fatty acid concentration decreased and the concentration of satrated fatty acids increased as fermentation proceeded. Satrated fatty acid concentration of the yeast cells correlated positively with high specific fermentation rates dring the later stages of fermentation in contrast to the belief that flid membranes with high nsatrated fatty acid concentrations enhance ethanol tolerance and, therefore, fermentation performance. The cell membrane Saccharomyces cerevisiae encloses the cytoplasm and selectively allows ntrients to enter and metabolic prodcts to leave the cell. Frthermore, cell wall components are deposited on this matrix dring growth of the cell It, therefore, affects growth and fermentation and is ltimately of interest to brewers and wine makers. Lipids are an integral part of celllar membranes and conseqently changes in lipid composition are associated with changes in membrane characteristics sch as flidity and permeability (Rattray, Schibeci & Kidby, 1975), which have been reported to affect ethanol tolerance of yeast (Thomas, Hossack & Rose, 1978). Factors which affect the lipid composition of micro-organisms are growth rate (Brown & Rose, 1969a), medim composition (Brown & Rose, 1969b), growth temperatre (McMrrogh & Rose, 1971), dissolved-oxygen tension (Brown & Rose, 1969a) and glcose concentration (Johnson, Nelson & Brown, 1972). Lower growth temperatres for example increase the proportion of nsatrated lipids in yeast cells (Pfisterer, Hancock & Garrison, 1977). However, according to McMrrogh & Rose (1971), temperatre variation does not significantly alter the fatty acid composition of the early exponential phase cells, bt profondly affects the synthesis of these componds dring the later exponential phase. Becase nsatrated fats melt at lower temperatres than their satrated conterparts, the change in composition related to growth temperatres has been described as an antifreeze effect; the increased proportion of nsatrated fats ensres a more flid membrane strctre at lower growth temperatres which presmably helps maintain membrane fnctions (patel & Lewis, 1982). Fatty acids are strctrally and fnctionally important and affect cell membrane fnction(s) to a great extent (Andreasen & Stier, 1954). The natre of the fatty acids (incorporated into phospholipids) affects the fnctions of certain proteins in the membrane (Pfisterer et al., 1977). The fnctional role of lipids, their fatty acid composition and degree of n satration, are regarded as major factors in the transport effects of membranes and the realization of fermentative activities in micro-organisms (Sinensky, 1971). Unsatrated fatty acids (UFA) in yeasts and other ecaryotic organisms are prodced by direct desatration of a pre-existing satrated fatty acid by a desatrase system which reqires moleclar oxygen (Bloch et ai., 1961). This desatrase enzyme system, however, becomes inactive when yeast cells grow in a medim containing nsatrated fatty acids (Radler, 1978). This stdy was ndertaken to determine the effects of temperatre dring fermentation on nsatrated and satrated fatty acid concentrations in for wine yeast strains and to establish whether there is a relationship between nsatrated or satrated fatty acid concentration and specific fermentation rate. MATRIALS AND MTHODS Fermentations: The yeast strains, media and methods described by Venter et al. (1989) were sed for the analyses and standardisation of yeast dry mass vales to correspond with 12,25,5,75, 1 gil and 125 gil sgar fermented (fermentation stages 1-6). Fermentation rates were previosly determinedbyventeretal. (1989) for the intervals -24, 13-37,38-62,63-87, gil and gil sgar fermented. These vales were considered to indicate the activity of the yeasts at stages 1-6, respectively, and were sed to determine specific fermentation rates (SFR) as gil sgar fermented/day/loo mg yeast cells. Mean satration index (Sl) was also determined for all treatments sing the following formla: SI = SFA UFA Gas chromatographic analysis of fatty acids: The method of Moss, Shinoda & Samels (1982) was modified and sed for fatty acid analysis. To 3 ml yeast sspension (5 mg to 2 mg yeast dry weight) in a 15 ml glass bottle was added 4 ml of methanol AR and,7 g NaOH, as well as 2 Acknowledgement: The athors wish to thank Professor M.A. Loos for a critical review of the manscript *Part of MSc. thesis presented at the University of Stellenbosch. Promotor: Prof. H.l.l. van Vren S. Afr. J. nol. Vitic., Vol. 1 No

2 ).II of a 5 mg/ml soltion of pentadecanoic acid (Sigma, St. Lois, USA) as internal standard. The sspension was flshed with nitrogen for 3 seconds to remove all oxygen and prevent peroxide formation. The bottle was sealed with a teflon-lined screw cap and placed in a boiling water bath for 1 hor. The saponified material was cooled for 15 mintes, and the ph adjsted to 2, by the addition of 4 ml 6 M HCL. The free fatty acids were then methylated by the addition of 4 ml of 14% boron trifloride-methanol complex reagent (BDH Chemicals, LTD, Poole, ngland) and heated for 15 mintes in a water bath at 8-85 C. The mixtre was cooled to room temperatre and methy I esters extracted by shaking with two sccessive 1 mlportionsofa 1:1 diethylether/hexane. The ether/ hexane layers were combined in a 5 ml beaker and the volme was redced to approximately 2 ml nder a gentle stream of nitrogen gas. The mixtre was sbseqently dried with MgSO 4 and transferred to a Reacti -vial where the volme was frther redced to,1 m!. The final volme was either analysed by gaschromatograph (GC) or stored at -2 C in a 2 ml screwcapped glass vial n analysis. The fatty acid methyl esters in the samples were analysed on a Varian 37 gas chromatograph eqipped with a flame ionization detector, sing a fsed silica capillary colmn (5 m x,3 mm LD.) with OV-Il as stationary phase (Hewlett Packard Corp., Avondale, Pa.). The injector temperatre was maintained at 26 C. Helim was sed as carrier gas at a flow rate of 1,39 ml/min. A 2).I1 sample which was split 5: 1 to yield a,4 ).II sample on the colmn, was injected. The colmn temperatre was controlled at 12 C for 3 mintes, then programmed at 5 C/min to 26 C, where it was maintained for 22 mintes before being cooled to the initial temperatre forthe next sample. The fatty acids from yeasts were identified by comparing the GC retention times of their methyl esters to those of athentic standards (Altech Associated Inc., Deerfield, USA). Peak areas were compared with that of the methyl ester of the pentadecanoic acid internal standard, and fatty acid concentrations determined by a Hewlett-Packard series laboratory system. The graphical method described by Venter et al. (1989) was sed for the standardisation of nsatrated and satrated fatty acid concentrations to correspond with the concentrations of sgar fermented at fermentation stages 1 to 6. Statistical analysis: A completely randomised experimental design was employed. Three measrements were taken at each of the six fermentation stages. Conseqently an analysis, taking accont of the potential correlations between measrements at the varios stages, was performed. The analysis was performed sing the RPATD command of the SAS procedre GLM and the POLYNOMIAL transformation keyword for fermentation stages. The PRINT option was sed to perform a sphericity test, as described in the SAS User's Gide (Anon., 1985). Separate analyses were performed for nsatrated fatty acids, viz. C14:1 +CI6:1 +CI8:1 and satrated fatty acids, viz. C8: + ClO:O + C12: + C14: + C16: + C18: and fermentation rates (Venter et at., 1989). A standard VORl factorial statistical software package was sed to test significant differences among treatment means (identical fermentation stages). The same program was sed to determine correlation coefficients. Fatty acid concentrations in wine yeasts 45 S. Afr. J. nol. Vitic., Vol. 1 No RSULTS AND DISCUSSION For all three variables (UFA concentration, SFA concentration and fermentation rate) strain x tempei:atre x stage interactions were indicated by the mltivariate test (the sphericity test was rejected in all three cases) and separate analyses were therefore performed for each yeast strain. The approximate significance levels (SL) for UFA and SFA concentrations were respectively,7532,,18 (W 452),,5445,,2542 (W 5),,5496,,1182 (W 14), and,8492,,259 (W 372). In the UFA concentration analyses by strain, the sphericity tests were not rejected, bt for the SF A concentration analyses, the sphericity test was rejected only in the case of W 452. In the case offermentation rate only the W 5 sphericity test was not rejected (SL =,1546). The analysis of fermentation rate by yeast strain showed stage x temperatre (qadratic) interaction, althogh there was a general tendency sggesting that the natre of the interaction cold be simplified to stage x temperatre (linear). Venter et al. (1989) showed that each yeast strain had different fermentation rate patterns, with higher specific fermentation rates at the higher temperatres. The rate of the decrease in UF A concentration (Fig. 1) and the variation in SFA concentration (Fig. 2) per 1 mg cells from the first (12 gil sgar fermented) to the sixth stage (125 gil sgar fermented) differed significantly between temperatres for all for yeast stains (data not shown). The UFA and SFA contents of yeast cells are considered to be individal characteristics of strains as indicated by the temperatre x strain x stage interaction. The decline in the UF A concentration of the yeast cells as the fermentations progressed can be ascribed to growth of the cells nder non-aerated conditions in the absence of exogenos sterols and nsatrated fatty acids (Venter et at., 1989). Occasional high vales at advanced fermentation stages cold be ascribed to experimental errors. The yeast dry mass increased considerably over the first three stages, bt little or no yeast growth was observed over the final three stages (Venter et at., 1989). The larger decreases in UFA concentration with yeast strains W 452 and W 5 than with strains W 14 and W 372 from the first to the second fermentation stage (Fig. 1) were the reslt of their larger increases in dry mass dring these stages. (Venter et at., 1989). The overall tendency oflower UF A concentrations at higher temperatres cold therefore be ascribed to a corresponding higher growth rate of yeast cells, reslting in a diltion of UF A as these componds are not prodced nder anaerobic conditions (Fig. 1). Yeast strain W 14 had the highest specific fermentation rates (SFR) at the first and second fermentation stages at 15 C and 2 C, followed by W 452, W 5 and W 372 (Venter et al., 1989). However, at the sixth fermentation stage yeast strain W 5 had the highest SFR of the for yeast strains at all three temperatres. The UFA concentrations showed a similar trend at 1 C with W 5 being higher than W 452 and W 372 bt at 15 C and 2 C no significant differences cold be detected (Fig. 1 & Table 1). At this (the sixth) stage SFR vales of W 452 and W 5 tended to be higher than

3 46 Fatty acid concentrations in wine yeasts -OJ ; LBO : g W 452 re. c nted. sgar fe rme W 5 -OJ ~ : 1.2 c Tempera tre. C sgar -OJ U 2.4 o --; BO W 372 ~ ~ ; 1.BO g re, c nted, fe rme gil 1. 2 : g re, c nted. sgar fe rme FIGUR 1 Unsatrated fatty acid concentration in cells of Saccharomyces cerevisiae strains W 452, W 5, W 14 and W 372 dring fennentation at three temperatres. that of W 14 and W 452 at 1 C and 15 C, albeit not always significant (Table 2). However, at 2 C yeast strains W 372 had the highest cell UF A concentrations (Fig. 1 & Table 1), bt the lowest SFR (Venter et al., 1989), whereas W 452 and W 5 had lower UFA concentrations with the highest SFR. Statistical analysis showed that a negative correlation (r = -,5) was fond between the UF A concentration and the SFR dring the sixth fennentation stage. Ths, a high UFA content alone does not ensre a high fennentation rate; it rather seems that a low UFA concentration correlates well with high SFR at 2 e. The SF A concentrations of strain W 14 increased as the fennentation proceeded (Fig. 2). However, in the case of strains W 452 and W 5 the SFA concentrations decreased considerably from the first to the second fennentation stage, whereafter they increased. The final SFA concentrations of the latter two strains were considerably higher than those of W 14 and W 372 (Table 2). TABL 1 Mean nsatrated fatty acid concentration in for yeast strains at three temperatres dring the sixth fennentation stage. Temp. (C) Unsatrated fatty acid concentration (mg/loo mg) W 452 W 5 W 14 W O,91b l,51a 15 O,89b O,85b 2 O,68b O,69b l,7a O,98b O,61b O,98b O,75b O,84b Means, within rows, followed by the same letter do not differ significantly (P.~O,5) S. Afr. J. nol. Vitic., Vol. 1 No

4 Fatty acid concentrations in wine yeasts 47 W 452 W5 ;'! ;'! n; ;:! ~ l~slo ~ o. i1 o. gil Il n ted. n>,d. 9 sgar fe rme f erme - r sgar Ure, c e, c ;'! W14 ;'! n; 4. n; 4. U ;:! 3..::: : 2. C <t <t - O. - o l~slo Uf'e. c sgar Ternperat llre, c sgar FIGUR 2 Satrated fatty acid concentration in cells of Saccharomyces cerevisiae strains W 452, W 5, W 14 and W 372 dring fermentation at three temperatres. When the SFA concentrations shown in Fig. 2 were compared with the SFR vales of Venter et at. (1989), it was fond that the former did not affect the latter dring fermentation. However, at the final stage a positive correlation (r =,87) was fond between the SF A concentrations and SFR of the for yeast strains. These reslts confirmed those of Nagar-Legmann & Margalith (1987) who showed that a fast-fermenting strain of Sacch. cerevisiae had higher SFA and free sterol concentrations than Sacch. mellis T 453 [reclassified as Zygo- saccharomyces cerevisiae (Von Arx, 1978)], which was a slow fermentor. Otogro et at. (1981) indicated that the membranes of Sacch. cerevisiae reqire certain amonts of SFA and that membrane fnctions (energy metabolism, transport, etc.) are impaired by SF A starvation. The importance of SF A in the membrane shold therefore not be nderestimated. Frthermore, it was obvios that higher satration index (SI) vales were generally obtained at the higher temperatres and these vales generally increased frther as the fermenta- TABL 2 Mean satrated fatty acid concentration in for yeast strains at three temperatres dring the sixth fermentation stage. Temp. (C) Satrated fatty acid concentration (mg/joo mg) W 452 W 5 W 14 W ,52ab 2,62a 15 2,9a 3,a 2 3,16a 3,92a 1,23b 1,97a 2,99a 1,38ab 1,96a 1,43b Means, within rows, followed by the same letter do not differ significantly (P.:::;.,5) S. Afr. J. nol. Vitic., Vol. 1 No

5 48 Fatty acid concentrations in wine yeasts TABL 3 The mean satration index for for yeast strains at six fermentation stages at three temperatres. Sgar fermented W 452 (gil) 1 C 15 C Mean satration index' W 5 W 14 W C 1 C 15 C 2 C 1 C 15 C 2 C 1 C 15 C 2 C 12,88 1,16 25,8 1,36 5,99 1, ,32 1,95 1 1,35 2, ,67 3,26 1,2 1,31 1,23 1,,89 1,1 2,5 1,27 1,45 2,88 1,5 2,52 3,68 1,4 2,32 4,65 1,73 3,49 1,4,77,99,88,84,95,92 1,52,85,94 1,51 1,92 1,1 1,14 1,94 i I 1,26 1,19 1,6 1,51 1,56 1,72 2,12 1,25 1,5 1,83 i 1,48 2,1 2,39 i 3,86 1,2 1,6 2,8 I 1,41 2,35 3,3 I 5,68 1,15 2,2 2,3 1,41 2,61 4,4! amean satration index = Satrated fatty acid conc. Unsatrated fatty acid conc. tion progressed (Table 3). It also appeared as if the cell membranes of yeast strains W 452 and W 5 at the sixth stage were less flid (relatively higher SFA concentrations) than the membranes of strains W 14 and W 372. Or reslts sggest that high membrane flidity is incompatible with high SFR dring the later stages offermentation. It has been sggested that high membrane flidity may be involved in the entrapment of ethanol molecles within the hydrophobic area of the membrane, interfering with its free movement ot of the cell (Legmann & Margalith, 1986). At the sixth fermentation stage the yeast cells contained relatively low concentrations of ergosterol (Venter et al., 1989) and UFA (Fig. 1). Watson & Rose (198) postlated a mechanism whereby membrane flidity cold be balanced throgh synthesis of phosphatidylinositol and phosphatidylserine containing a higher proportion of satrated fatty-acyl resides, strategically located in the membrane and thereby ANDRASN, A.A., & STIR, TJ.B., Anaerobic ntrition of Saccharomyces cerevisiae. n. Unsatrated fatty acid reqirement for growth in a defmed medim. 1. Cell. Compo Physiol. 43, ANONYMOUS, SAS User's Gide: Statistics, Version 5 dition, SAS Institt Inc., Cary, NC. BLOCH, K. BARONOWSKY, P., GOLDFIN, H., LNNARZ, W.J., LIGHT, R., NORRIS, A.T. & SCHURBRANDT,G., Biosynthesis and metabolism of nsatrated fatty acids. Fed. Proc. Fed. Am. Soc. xp. BioI. 2, BROWN, C.M. & ROS, A.H., 1969a. ffects of temperatre on composition and cell volme of Candida titis. 1. Bacteriol. 97, BROWN, C.M. & ROS, A.H., 1969b. Fatty acid composition of Candida is as affected by growth temperatre and dissolved-oxygen tension. 1. Bacteriol. 99, JOHNSON, B.S., NLSON, CJ. & BROWN, C.M., nf1ence of glcose concentration on the physiology and lipid composition of some yeasts. Antonie van Leewenhoek,l. Microbiol. 38, LGMANN, R. & MARGAL1TH, P., thanol formation by hybrid yeasts. r.l. Appl. Microbiol. Biotechnol. 23, McMURROUGH, 1. & ROS, A.H., ffects of temperatre variation on the fatty acid composition of Candida titis. 1. Bacteriol. 17, MOSS, C.W., SHINODA, T. & SAMULS,J.W.,1982. Determination of celllar fatty acid compositions of varios yeasts by gas-liqid chromatography. 1. Clin. Microbiol. 16, NAGAR-LGMANN, R. & MARGALITH, P., A comparative stdy of the lipid composition of yeasts with different fermentative capacities. r. 1. Appl. Microbioi. Biotechnol. 26, LITRATUR CITD maintaining a degree of rigidity ;n that domain. CONCLUSIONS S. Afr. J. nol. Vitic., Vol. 1 No Fermentation temperatres affected the fermentation rate, UF A and SF A concentrations of different yeast strains to a great extent. High specific fermentation rates at the later stages of fermentation were correlated with high SFA concentrations. It appears to be in contrast with the findings of other workers i.e. that flid membranes containing relatively high concentrations ofufa, might interfere with the movement of ethanol ot of the yeast cell. A high intracelllar concentration of ethanol will impair fermentation capabilities of yeast. On the other hand, reslts from this stdy in synthetic media indicate that high concentrations of SF A might well render yeast cells more tolerant to ethanol than their nsatrated conterparts, enhancing fermentation rates in the presence of ethanol dring the later stages of fermentation. OTOGURO, K., A WA Y A, J., TANAKA, H. & OMURA, S., Satrated fatty acidstarved cells of Saccharomyces cerevisiae grown in the presence of cerlenin and oleic acid. 1. Biochem. 89, PATL, P.C. & LWIS, M.J., nf1ence of growth temperatre on the fatty acid composition of the cytoplasmic membrane of a larger yeast. 1. Am. Soc. Brew. Chem.4, PFISTRR,., HANCOCK, I. & GARRISON, I., ffects of fermentation environments on yeast lipid synthesis. 1. Am. Soc. Brew. Chem. 35, RADLR, F., Promotors of the yeast anaerobic growth. Ann. Technol. Agric. 27, RATTRAY, J.B.M., SCHIBCI, A. & KIDBY, O.K., Lipids of yeasts. Bacteriol. Rev. 39, SINNSKY, M., re control of phospholipid biosynthesis in scherichia coli. 1. Bacteriol. 16, THOMAS, D.S., HOSSACK, J.A. & ROS, A.H., Plasma-membrane lipid composition and ethanol tolerance in Saccharomyces cerevisiae. Arch. Microbiol. 117, VON ARX, J.A., Centraalbrea voor Schimmelcltres. Progress Report Verh. Kon. Ned. Akad. Wetensch., Afd. Natrk., 2e Reeks, 71, VNTR,J.J., V ANVUURN,HJJ., TROMP,A. &RANDALL,J.H., Relationship between ergosterol concentration in wine yeasts and sgar fermentation at different temperatres S. Afr.l. nol. Vi/ic. 1, WATSON, K. & ROS, A.H., 198. Fatty-acyl composition of the lipids of Saccharomyces cerevisiae grown aerobically or anaerobically in media containing different fatty acids. 1. Gen. Microbiol. 117,