SULFITE AND SULFIDE FORMATION -- A REVIEW

Transcription

1 SULFITE AND SULFIDE DURING WINEIVIAKING FORMATION -- A REVIEW R. ESCHENBRUCH Senior Research Officer, Oenological and Viticultural Research Institute, Stellenbosch, South Africa. address: Ruakura Agricultural Research Centre, Private Bag, Hamilton, New Zealand. The author thanks Dr. D. E. G. Sheat for constructive review and corrections of the manuscript. Accepted for publication July 8, Present ABSTRACT Present knowledge of sulfite and sulfide formation by wine yeasts during the fermentation of grape juice is reviewed. Possible sources of sulfur, namely sulfate, sulfite, elemental sulfur and sulfurcontaining amino acids, are discussed. Factors in- fluencing sulfide as well as sulfite formation are emphasized. Low-sulfite and high-sulfite-forming yeast strains differ in their metabolism as well as in some aspects of regulation. Introduction: The ability of yeasts to form sulfur dioxide has been known since the end of the last century. This, however, has for many decades never been seriously considered with regard to winemaking even though sulfite is the most commonly used microbiostatical and antioxidative additive facilitating the production of wine. As awareness grew of the importance of the quality and purity of foodstuffs, laws and regulations limited the use of addirives, preservatives, etc. in such products. In spite of many efforts, no real substitute for sulfite has been found so far. Medical and health experts, however, keep warning against excessive consumption of sulfite, used not only in wine but also in many other foodstuffs (25, 42). A special study group of the Office International de la Vigne et du Vin aims at reducing the addition of S02 to grape juice and wine (21, 30, 56). The formation of hydrogen sulfide by yeasts during the fermentation of grape juice is a problem as old as the process of winemaking. Hydrogen sulfide, which smells and tastes most unpleasant, is one of the highly undesirable metabolites of alcoholic fermentations. Although much information has been gathered during the last 15 years the problem of its control is basically unsolved. Only a few significant references from numerous reports show its general importance (23, 28, 37, 47). This article reviews sulfite and sulfide formation during winemaking. 157 Sulfur sources of sulfite and sulfide: Sulfur, which is essential to yeast growth, is naturally available in abundance as sulfate in grape juice. Yeasts reduce this sulfate to sulfide via adenosine-5'-phosphosulfate (APS), 3'-phosphoadenosine-5'-phosphosulfate (PAPS), and sulfite (40). This is illustrated in figure 1, which is a simplification of the metabolic pathway of sulfur in yeasts. Methionine ,Protein t \ ATP ATP NADPH I "~ Mercaptan $ ~ I SO~ = -oaps ~ PAPS --~SO. =, S = Cysteine ~ Pyruvate, NHa Figure 1. Simplified pathway of sulfur in wine yeasts.,protein Although sulfate is the only substrate for any sulfite formation, it is not the only source of sulfide; sulfite can be used as easily (17, 23, 28, 37, 41). Furthermore, elemental sulfur, applied as fungicide in vineyards and also derived from insufficiently burned sulfur candles, provides another source (1, 8, 22, 23, 28, 37, 41). Free sulfur is said to be nonenzymically converted to H2S with reducing compounds provided by the yeast (47). As seen in figure 1, cysteine can also be a precursor of H~S since yeasts can decompose it to

2 H2S, pyruvate, and NH3 (5). Methionine, however, the other major sulfur-containing amino acid present in musts, inhibits H2S formation (17, 23, 47). Factors influencing sulfide formation: Apart from the compounds of sulfur metabolism there are other factors influencing H2S production indirectly. One of these is the yeast strain. The amount of H2S produced under defined conditions is characteristic of the strain. Selection of suitable strains has so far been the most successful way of limiting excessive H2S formation. Such pure-culture yeasts have been long recommended to winemakers (e.g., 7, 8, 36, 41). All aspects of fermentative yeast growth and metabolism are involved. The greater the amount of growth the more sulfide is formed. For a detailed discussion see (47). Reduction of sulfate and sulfite requires production of ATP and NADPH (Figure 1), and biosyntheses of organic sulfur compounds, such as cysteine, methionine, coenzyme A, glutathione, thiamine, and biotin, are preceded by sulfide formation. Pantothenate and pyridoxine deficiencies in many strains of Saccharomyces cerevisiae result in low levels of methionine production in the cells, with consequent H2S formation (23, 28, 45, 46). AI: though pantothenate, pyridoxine, thiamine, and biotin are not normally lacking in grape musts (33), they may sometimes be limiting when various preparations and treatments of grape juice are considered, e.g., heat treatments. Whether a new yeast strain is pantothenate- or pyridoxine-deficient should be checked before it is released for practical winemaking. Certain metal ions stimulate the formation of sulfide by yeast. The conclusion that increased H2S production during fermentation of grape juice could not be related to copper, zinc, or tin ions (37) has not been confirmed (11, 13, 18). When used in vineyards up to six weeks before harvest, fungicides such as copper oxychloride and Bordeaux can cause in unfermented musts a two-to-four-fold increase in copper content which considerably stimulates hydrogen sulfide formation during fermentation (18). Similar effects have been reported for zinc and tin ions in brewing (12, 27, 43, 47). More and more, zinc- and manganese-containing fungicides are being used in vineyards, but no information has yet come to our knowledge on the increased concentration of these metals in grape juice or their influence on H2S formation. With the South African grape cultivar 'Steen' (var. V. vinifera) the H2S formation was said to be inversely proportional to the nitrogen content of the must (44). Extensive tests over three successive vintages with musts of the same variety but grown in different viticultural areas and employing five different yeast strains could not confirm these findings (20). One intriguing result was, however, that within each vintage one must always gave a higher SULFURS FORMED IN WINEMAKINGm158 H.oS production with all five yeasts, though this particular must was a different one in each season (20). The reason is not yet clear. Furthermore, in all three years the must with the lowest amino acid content never gave the highest H2S formation (20). It thus appears that it is neither a low nor a high amino acid content that determines the potential of a must for H2S formation. Recent reviews of H2S production (46, 47, 49) concluded that unless there is a relative deficiency of methionine most yeasts form only little free H2S from sulfate, and that methionine deficiency can arise if the medium contains insufficient pantothenate or vitamin BG or contains an excess of certain amino acids; aspartic acid, glutamic acid, glycine, homoserine, histidine, lysine, ornithine, serine, and threonine. The free amino acid contents of the 'Steen' must samples indicated no excess of those amino acids, but methionine content was very low, varying from 1 to 20 mg per liter (20). All samples, including those from higher H2S-producing musts, seemed deficient in methionine. Responsibility for H2S production therefore could not be due solely to a deficiency of this amino acid. Cysteine was also measured in the "Steen' must samples but only traces were detected (19). Similar results have been reported for some Spanish musts (24, 34). Neither the formation of hydrogen sulfide from breakdown of cysteine nor the production of off-flavors from decompositions of methionine seemed pertinent with those musts. Autolysis is the self-digestion of yeast cells, which probably occurs throughout the fermentation, but it is most marked toward the end of this process and particularly if wine is left on the lees for too long. Under these conditions a strong and often incurable H2S or mercaptan smell may develop. Very little H2S is detectable, but other sulfhydryl compounds are formed which may account for the off-flavor (51). Four proteolytic enzyme fractions have been partially characterized (26) ; these appear to cause the breakdown of the intracellular proteins. If proreins are decomposed, methionine as well as cysteine is released, and their metabolism could cause offflavors. However, too little is known about this overall problem and most recommendations to the winemaker stem from empirical observations. Factors influencing sulfite formation,: The formation of sulfite by yeasts during fermentation has been investigated intensively since 1965 because of its potential legal consequences (55). This formation has been reported from various wine-producing countries (19, 29, 32, 35, 38, 50, 52), and yeast strains have been shown to differ in their ability to form sulfite. If sufficient numbers of strains of S. cerevisiae are tested, most form between 10 and 30 mg S02 per liter. Under comparable conditions, some strains produce ess than I0 mg, whereas others form more than 100 mg per liter (8, 31, 53, 54, 55). The latter have been called "S02-forming strains"

3 159--SULFURS FORMED IN WINEMAKING to distinguish them from the "normal yeasts", which form sulfite but to a far lesser extent. It is, however, more exact to use the terms "low" or "high" sulfite-forming strains. But low-sulfite-forming strains sometimes do produce more sulfite, and up to 39 mg S02 per liter was measured in two different musts fermented by two pure-culture yeasts (19). In a simple defined medium the same yeasts synthesized about 60 mg per liter (15). The regulatory mechanisms of these changes are not yet understood. In 1968 it was postulated that the high-sulfiteforming yeasts were defective mutants, unable, on account of a genetic block, to reduce sulfite to sulfide (Figure 1), so that sulfite would accumulate (8,9). Subsequently, however, it was demonstrated that these high-sulfite-forming yeasts form amounts of H2S similar to those of low-sulfite-forming strains (14). Similar results were also reported for one high-sulfite-forming strain from America (1). It was further shown that yeasts forming little sulfite take up comparatively little sulfate. On the other hand, high-sulfite-forming strains also have a correspondingly high sulfate uptake (16). Finally, using labeled sulfate and sulfite, it was shown that the amount of H2S produced from sulfate and sulfite by high-sulfite-forming yeasts was similar to that produced by low-sulfite-forming strains (17). The sulfate concentration of grape juice varies (19), but does not influence sulfite formation by low-sulfite-forming strains since sulfate is always abundant. With high-sulfite-forming strains the situation is different. Increasing additions of sulfate cause increased sulfite production up to saturation, such that the formation of more than 500 mg sulfite per liter has been reported (32, 55). Regulation of sulfite and sulfide formation: Figure 1 shows that methionine and cysteine are the main end products of the sulfate reduction sequence. Since all amino acids regulate their own synthesis, the rate of sulfate reduction is influenced by the need of the yeast cell to synthesize these amino acids. In other words, the higher the concentration of methionine and cysteine in the grape must, the smaller the amount that has to be formed, the lower the rate of sulfate reduction and consequently the lower the formation of sulfite and sulfide. Various tests, including some with radioactive sulfur, demonstrated that the addition of increasing concentrations of methionine or cysteine to grape juice reduces or completely suppresses the uptake of sulfate and the formation of sulfite and sulfide by low- as well as high-sulfite-forming strains (15, 16, 17). Similar results have been reported for brewing yeasts (2, 23, 48). It has been emphasized that increasing pressure during grape pressing increases the free amino acid content of the must (4), and methionine concentrations up to 90 mg per liter have been reported (10, 30). Such levels might influence sulfite and sulfide formation, but to date too little is known about the extent to which the amino acid content of musts is affected by different varieties, vintages, climatic conditions, soil types, and treatments with technical enzymes, e.g., proteases. The addition of these amino acids to grape juice has, however, no practical significance. As pointed out earlier in this article, eysteine can be converted to H2S, pyruvate, and NH3. Methionine can be metabolized to 7-methyl-mereaptopropanol, its aldehyde and ester (3, 48), compounds which are as detrimental to wine as is H2S. Sulfite is an intermediate of sulfate reduction, and intermediate products of biochemical sequences can regulate their formation as well as their turnover. Since few grape musts are fermented without the addition of sulfur dioxide, the additive should exert a distinct influence on the reduction of sulfate. Furthermore, it is well known that up to 50% of added S02 disappears during fermentation. Whether it is reduced (41) or oxidized (6) is not well established. When labeled sulfite at 150 mg SO., per liter was added prior to fermentation, 30~ of the label appeared in sulfate and about 10% in reduction compounds (17). The same pattern pertains, though less so, to two high-sulfite-forming strains (18). Consequently, part of the added S02 is oxidized and part reduced. Quite interesting, though not yet understood, is the finding that when methionine and cysteine (at 500 mg/liter each) are added, the formation of sulfate from labeled sulfite doubled and this oxidation in the high-sulfite-forming strains increased by four-to-five-fold (17). Not only the magnitude of the increase in oxidation but also the differences between strains argues against a purely chemical reaction. Furthermore, information on regulation of the sulfate-reduction sequence was obtained from the experiments with labeled sulfur. The conversion of sulfate to sulfite and sulfide was almost completely suppressed by the addition of sulfite (17). Sulfite, therefore, effectively controls its own formation in low-sulfite-forming strains. The sulfate turnover of high-sulfite-forming strains was, however, not influenced at all by exogenous S02 (17). These results offer a new approach toward understanding of excessive sulfite formation by some strains. It would appear that in these strains the enzymes involved in the formation of sulfite are no longer controlled and regulated by the intermediate sulfite. This view is supported by data demonstrating that increasing additions of sulfate to must cause increasing sulfite formation by high-sulfiteforming strains but not by low-sulfite-forming yeasts (32, 55). The validity of this concept needs confirmation by direct determination of the activity of the enzymes involved.

4 SULFURS FORMED IN WlNEMAKING--160 Furthermore, these results could explain why low-sulfite-forming strains seldom exhibit sulfite formation under practical conditions during winemaking. Added sulfite not only prevents the formation of sulfite from sulfate but is itself also oxidized and reduced. Notably, distilling wines in South Africa are being produced without S02 addition. Under such conditions sulfite formation of pureculture yeasts is encountered; occasionally the legal limit of 25 mg per liter is exceeded (19). 8. Dittrich, H. H. and Th. Staudenmayer. SO2-Biidung, BSskerbildung und BSckserbeseitigung. AIIgem. Deutsche Weinzeitung 104:707-9 (1968). 9. Dittrich, H. H. and Th. Staudenmayer. 0ber die Zusammenh&nge zwischen der Sulfit-Bildung und der Schwefelwasserstoff-Bildung bei Saccharomyces cerevisiae. Zentribl. Bakt. Abt. I1.124:113-8 (1970). 10. Dittrich, H. H., E. Leidenfrost and W. Tepe. 0bertragt sich eine hohe NO3-Dfingung der Reben auf den NO3-Spiegel im Most? Wein-Wiss. 25:130-2 (1970). Conclusions: Sulfite and sulfide formation are aspects of winemaking of considerable economic importance. Their production can be partially controlled. The problem of the formation of both can currently be overcome by employing strictly defined pure-culture yeast strains. Considering, however, the world-wide trend toward lower S02 levels in wines, does the ultimate aspiration for S02-free wines then become an impossibility? Because the production of sulfite and sulfide is an integral part of yeast metabolism it is doubtful whether complete control is possible. Isolation of more suitable yeast strains, prevention of excess of certain metal ions in the must, and elimination of elemental sulfur can reduce the amount of H2S formed. Further information is required on the influence of the composition of varietal musts, new cellar technologies, and intracellular sulfur metabolism of yeasts. How are undesirable sulfur compounds produced in the lees toward the end of a fermentation? Are these the compounds of yeast catabolism or autolysis? Answers to these questions will undoubtedly benefit the quality of wines. LITERATURE CITED 1. Acree, T. E., E. Sonoff and D. F. Splittstoesser. Effect of yeast strain and type of sulfur compound on hydrogen sulfide production. Am. J. Enol. Viticult. 23:6-9 (1972). 2. Anderson, A. J., G. A. Howard and J. S. Hough. The sulphur metabolism of brewing yeasts and spoilage bacteria. Eur. Brew. Conv. Proc. 1971: Bfrwald, G. and D. Kliem. Ein Beitrag zum Methioninstoffwechsel der Hefe (S. cerevisiae). Chem. Mikrobiol. Technol. Lebensm. 1:27-32 (1971). 4. Bergner, K. G. and H. E. Hailer. Das Verhalten der freien Aminos&uren von Weissweinen im Verlauf der G&rung, bei Ausbau, Lagerung und Umg~irung. Mitt. Klosterneuberg 19: (1969). 5. Dagley, S. and D. E. Nicholson. An introduction to metabolic pathways. Blackwell Scientific Publications, Oxford and Edinburgh (1970). 6. Diemair, W., J. Koch and D. Hess. 0ber den Einfluss der schwefligen Saure und L-Ascorbins~iure bei der Weinbereitung.!. Mitt. Ober die Bindung der schwefligen Sfure an Acetaldehyd und Glucose. 7. Dittrich, H. H. Die Entstehung von "Schwefel"-bindenden und Schwefe-haltigen Hefe-Stoffwechselprodukten. Aiigem. Deutsche Weinfachzeitung 109:37-42 (1973). Amer. J. Enoh Viticult., Voh 25, No. 3, Dittrich, H. H. and Th. Staudenmayer. 0ber die ErhShung der H2S-Bildung wfhrend der G~irung durch Kupfer-lonen. Wein-Wiss. 27:250-3 (1972). 12. van Engel, E. L. Zinc, metalloenzyme activity and the growth of brewers yeast. A.S.B.C. Proc. 1969: Eschenbruch, R. The influence of fungicides on the formation of hydrogen sulphide (H2S) during the fermentation of grape juice. Wynboer No. 482:22-3 (1971). 14. Eschenbruch, R. Zur Substratabh&ngigkeit der H2S- und S02-Bildung bei Saccharomyces cerevisiae St~immen. Wein- Wiss.. 27:40-4 (1972). 15. Eschenbruch, R. Der Einfluss von Methionin and Cystein auf die SO2-Bildung einiger St&mme von Saccharomyces cerevisiae bei der Verg&rung von Traubenmost. Vitis 11:53-7 (1972). 16. Eschenbruch, R. Sulphate uptake and sulphite formation related to the methionine and/or cysteine content of grape must during the fermentation by strains of Saccharomyces cerevisiae. Vitis 11:222-7 (1972). 17. Eschenbruch, R., F. J. Haasbroek and J. F. de Villiers. On the metabolism of sulphate and sulphite during the fermentation of grape must by Saccharomyces cerevisiae.. Arch. Mikrobiol. 93: (1973). 18. Eschenbruch, R. and P. H. Kleynhans. The influence of copper-containing fungicides on the copper content of grape juice and on hydrogen sulphide formation. Vitis 12:320-4 (1974). 19. Eschenbruch, R. On the sulphite formation and suiphite consumption of wine yeasts. Wynboer No. 506:42-4 (1973). 20. Eschenbruch, R. On the hydrogen sulphide formation by wine yeasts. Wynboer No508: 8-12, 46 (1974). 21. Gomes, Marques J. V. Limitation de I'empioi de I'anhydride sulfureux compte tenu des besoins en mati~res nutritives des levures pendant la fermentation. Bull. O.I.V. 46: (1973). 22. Hernandez, R. Production of H2S by wine yeasts grown with several sulphur containing compounds. Semana Vitivinicola 19: (1964). (Chem. Abstr. 64: 13353, 1966). 23. Lawrence, W. C. and E. R. Cole. Yeast sulfur metabolism and the formation of hydrogen sulfide in brewery fermentations. Wallerst. Lab. Commun. 31: (1968). 24. Llaguno, C. Autoanalysis in oenology laboratories. Third Intern. Oenolog. Symposium, Cape Town 1972, paper No L0thi, H. R. Wie weit ifsst sich SO 2 in der Weinbereitung ersetzen? Third Intern. Oenolog. Symposium, Cape Town 1972, paper No Maddox, Io S. and J. S. Hough. Proteolytic enzymes and autolysing brewers' yeast. Proc. Europ. Brew. Conv. 12: (1969).

5 161--SULFURS FORMED IN WINEMAKING 27. Maddox, I. S. and J. S. Hough. Effect of zinc and cobalt on yeast growth and fermentation. J. Inst. Brew. 76:262-4 (1970). 28. Maw, G. A. The role of sulphur in yeast growth and in brewing. Wallerst. Lab. Commun. 28:49-70 (1965). 29. Mayer, K. and P. Pause. 0ber die Bildung von schwefliger S~iure und Schwefeiwasserstoff w~ihrend der Weinggtrung. Mitt. Lebensm.-Unters. u. Hygiene 59: (1968). 30. Mayer, K. and A. Dufour. Etude sur ia production de SO 2 par les levures au cours de la fermentation. Rapport du Groupe de Travail "Microbiologie du Vin", O.I.V. May Minarik, E. SO2-Bildung durch Sulfatreduktion bei verschiedenen Hefearten der Gattung Saccharomyces. Mitt. Klosterneuburg 22: (1972). 32. Minarik, E. Formation of sulphurous acid through reduction of sulphates during fermentation of must. Kvasny prumysl 18:104-7 (1972). 33. Ournac, A. Les vitamines du vin. Ann. Nutr. Alim. 24:B333- B365 (1970). 34. Polo, M. C. and C. Llaguno. Determinacion de componentes nitrogenados en mostos de uva y vino, y su influencia en los procesos de crianza de vinos. Simposio Intern. Technicon "Automazione per la Medicina e per I'lndustria", Roma, April Premuzic, D., T. Lovric, O. Safar and V. Jovic. Production of sulphur dioxide during fermentation of must as a result of the metabolism of some yeast strains and their effect on the colour of white wines. Kem. Ind. 21:9-20 (1972). (Chem. Abstr. 76:152021g, (1973). 36. Rankine, B. C. The importance of yeasts in determining the composition and quality of wines. Vitis 7:22-49 (1968). 37. Rankine, B. C. Nature, origin and prevention of hydrogen sulphide aroma in wines. J. Sci. Fd Agr. 14:79-91 (1963). 38. Rankine, B. C. and K. F. Pocock. Influence of yeast strain on binding of SO 2 in wines and on its formation during fermentation. J. Sci. Fd Agr. 20:104-9 (1969). 39. Rapp, A. and K. H. Reuther. Der Gehalt an freien Aminos~iuren in Traubenmosten von gesunden und edelfaulen Beeren verschiedener Rebsorten. Vitis 10:51-8 (1971). 40. Roy, A. B. and P. A. Trudinger. The biochemistry of inorganic compounds of sulphur. University Press, Cambridge (1970). 41. Schanderl, H. Die Mikrobiologie des Mostes und Weines. Verlag Eugen Ulmer, Stuttgart (1959). 42. Schroeter, L. C. Sulfur dioxide. Application in food, beverages and pharmaceutical. Pergamon Press, London (1966). 43. Thorn, J. A. Yeast autolysis and its effect on beer. Brewers' Digest, Oct 1971, (J. Inst. Brew. 78:351, 1972). 44. Vos, P. J. A. Korrelasies tussen gesondheidstoestand van Steendruiwe, chemiese samestelling van die mos en organoleptiese beoordeling van die wyn. Masterate Thesis, University of Stellenbosch (1966). 45. Wainwright, T. Hydrogen sulphide production by yeast under conditions of methionine, pantothenate or vitamine B 6 deficiency. J. gen. Microbiol. 61: (1970). 46. Wainwright, T. Wort composition and hydrogen sulfide formation. A.S.B.C. Proc. 1970: Wainwright, T. Production of H2S by yeasts: role of nutrients. J. Appl. Bact. 34: (1971). 48. Wainwright, T. Origin and control of undesirable sulphur compounds in beer. Proc. Europ. Brew. Conv. 1971: Wainwright, T. Effect of leucine on H2S production by yeast; relationship to methionine. Brauwiss. 24:334-8 (1971). 50. Weeks, C. Production of sulfur dioxide-binding compounds and of sulfur dioxide by two Saccharomyces yeasts. Am. J. Enol. Viticult. 20:32-9 (1969). 51. Wisser, K. Zum Auftreten von Sulfhydrylverbindungen bei der Autolyse yon Saccharomyces cerevisiae. Chem. Mikrobiol. Technol. Lebensm. 2:77-80 (1972). 52. W0rdig G. and H. Schlotter, SO2-Bildung in g&renden Traubenmosten. Z. Lebensm.-Unters. u. Forsch. 134:7-13 (1967). 53. Wfirdig, G. and H. A. Schlotter. SO2-Bildung durch Sulfatreduktion w&hrend der G,~rung. I. Versuche und Beobachtungen in der Praxis. Wein-Wiss. 23: (1968). 54. Wfirdig, G. and H. A. Schlotter, SO2-Bildung durch Sulfatreduktion w~ihrend der G~irung. II. Beeinflussung durch das Substrat und die G~irungsbedingungen. Wein Wiss. 25: (1970). 55. W0rdig, G. and H. A. Schlotter. 0ber das Vorkommen SO2-bildender Hefen im nat~jrlichen Hefegemisch des Traubenmostes. Deutsche Lebensm. Rdschau 67:86-91 (1971). 56. W(Jrdig, G. Formation de SO 2 par r~duction du sulfate pendant la fermentation. Rapport du Groupe de Travail "Microbiologie du Vin." O.I.V. April Amer. J. Enol. Viticult., Voi. 25, No. 3, 1974