[Advances in Applied Microbiology] Volume 12 || Flavor and Microorganisms

Flavor a n d Microorganisms

P. MARGALITH AND Y. SCHWARTZ Department of Food and Biotechnology, Technion -Israel Institute of Technology,

Haifa, Israel

I. General Introduct‘ .... ....

and Development of Flavor in Traditional Foods ............ A. Baked Products ....... B. Fermented Beverages ............................................ C. Dairy Products ...................................................... D. Pickles ................................................................. E. Oriental Foods ........

111. Concluding Remarks .................................................... References ..................................................................

11. The Contribution Pr 36

40 40 45 64 68 72 74 83

The spectacular increase in world population has led to a major effort, on a national and international basis, for the increased production of foods. Although substantial advances have been made in pi-oducing nutritionally valuable food ingredients of low cost, such as protein from microbial sources (Champagnat et al., 1963; Fiechter, 1967) and fishmeal concentrate, the acceptance of these foods in the needy developing countries where undernourishment is frequently encountered is far from satisfactory; one of the major disadvantages of these unconventional foods are their poor taste. Many workers have pointed out that after all the nutritional requirements of foods have been met with, the incorporation of suitable flavoring material into these “synthetic foods” have become imperative (Hornstein and Teranishi, 1967).

There have been spectacular advances in the characterization of flavor. Gas-liquid chromatography and mass spectroscopy have opened new fields in the analysis of flavor and flavoring materials. I t seems, however, that analytical devices and instrumentation are stil€ lagging behind the discriminatory power of human taste.

Microbiology has played a prominent role in the production of fermented foods and beverages. The production of alcohol has been considered the major contribution of yeasts, while that of the lactic acid bacteria was mainly concerned with the transformation of sugars into lactic acid. The contribution of microorganisms to the development of flavor was not as widely investigated and only fragmentary information is available. With the increase in knowledge of the chemical nature of flavor it was discovered that microorganisms not only

35

36 P. MARGALITH AND Y. SClIWARTZ

contribute to taste in fermented products, but may be also employed for the specific production of food additives with flavor-enhancing properties on an industrial scale (Kuninaka, 1966; Solms, 1967).

In this survey an attempt will be made to review the information available on the contribution of microorganisms to flavor as well as the fermentative production of some major flavor enhancing materials. However, it must be pointed out that the subject of microbial food spoilage is outside the scope of this review.

I . General Introduction

In the interest of the general reader, a brief introduction to the nature of flavor will be made. Moncrieff (1951) defines flavor as “a complex sensation comprising taste, odor, roughness or smoothness, hotness or coldness, and pungency or blandness.” Although taste and odor are generally considered as main attributes of flavor, it should be remembered that flavor is mainly the psychological attitude of the human senses, and not the inherent properties of any material (Horn- stein and Teranishi, 1967).

Taste is the sense peculiar to the mouth and tongue. Historically a large number of components of taste have been discerned, although today it is generally agreed that taste comprises four basic taste qualities, i.e., sweet, salty, sour, and bitter (von Skramilk, 1921). However, niodifications of these basic qualities are frequent; expres- sions such as metallic, alkaline, and fatty tastes may be encountered.

The sensation of taste is achieved by the sense buds distributed all over the tongue and some restricted areas in the buccal cavity. Their number is estimated to be between 9000 and 10,000. Taste buds are located in tiny papillae in the anterior part of the tongue, whereas in the posterior portion they line the sides of minute trenches. Each taste bud consists of about 10 to 15 taste cells from which very minute villi protrude. The taste buds are innervated b y nerve fibers arising from the subepithelial plexus. For a detailed description of the anatomy of these tissues the reader is referred to a histological text (e.g., Amerine et al., 1965). Although it was previously assumed that there exist different taste buds or receptor cells that respond to different taste stimuli according to the four basic taste qualities, this has been recently shown not to be the case. In fact, all taste buds respond to many of the taste principles, although quantitatively they are not alike. However, it is the integrated taste response of a multitude of taste buds that create the sensation of taste ( P f a b a n n , 1964; Beidler, 1966).

There have been many efforts to explain the mechanism by which

FLAVOR AND MICROORGANISMS 37

taste active materials cause the sensation of taste. Beidler (1966) has shown that the initial step in the taste stimulation is a weak adsorption of the active material onto the receptor site on the surface of the taste cell. This in turn causes a depolarization in the charge of the nerve fiber leading to the formation of a nerve impulse. The involvement of enzymatic reactions in the creation of the taste stimulus has been suggested by Duncan (1963).

One of the basic features in the evaluation of taste activity, is the determination of the threshold value of the specific material, i.e., the minimum detectable concentration of the test substance. For comparison the following threshold values of common taste materials will be shown below:

Material Threshold value (%)" ~~

Sodium chloride Hydrochloric acid Sucrose Quinine

0.25 0.007 0.5 0.00005

Howell, 1922.

Since similar taste qualities may be achieved by compounds of different chemical structure, a brief discussion of the correlation between structure and sensation is pertinent. Sourness is achieved only by acids, most of their activity being probably due to the dissociation of hydrogen ions (Amerine et al., 1965). The sensation of saltiness is produced by many low molecular salts, both cations and anions contributing to a different extent to this sensation. As the molecular weight of the salt increases, in addition to saltiness other qualities arise, primarily bitterness. Thus, the saltiness decreases in the order: KCI, KBr, KI, while the bitterness increases accordingly (Moncrieff, 1951).

While sourness and saltiness are produced by distinct groups of chemical compounds, bitterness and sweetness occur in a large variety of unrelated compounds. From the chemical point of view bitterness and sweetness seem to be very close. In fact there are a number of compounds which are structurally very close but cause distinct taste qualities. For example, 2-amino-4-nitropropoxybenzene is about 4000 times sweeter than sucrose, 4-amino-2-nitropropoxybenzene is tasteless, while 2, 4-dinitropropoxybenzene is bitter (Hornstein and Teranishi, 1967). Also, the taste stimulus is greatly affected by iso- meric changes. Thus, o-nitrobenzoic acid is very sweet, m, slightly

38 F. MARGALITH AND Y. SCHWARTZ

sweet, and p , is bitter (Moncrieff, 1951). On the other hand, completely different molecules can produce a similar taste stimulus. Sucrose, saccharine (o-sulfonbenzimide), and cyclamate (cyclohexyl sulfa- mate) are widely known examples.

OCH,CH,CH, OCH,CH&€& OCH,CH,CH,

@ NH2 o-” +NOZ

NO2 NH2 NO,

(2- Amino-l-nitro- (4-Amino-2-nitro- (2,4-Dinitro- propoxybenzene) propoxybenzene) propoxybeneena)

Recently, the effect of the optical configuration on the taste activity of a number of compounds has been studied. The L-isomer of glucose has been found not to be sweet but slightly salty (Boyd and Matsu- bara, 1962). The anomers of mannose show a different taste activity, a-D-mannose being very sweet while p-D-mannose is very bitter (Steinhardt et aZ., 1962). Different tastes have been attributed to the stereoisoniers of amino acids. For example, L-isoleucine is bitter, whereas the D-isomer is sweet (Berg, 1953).

The threshold value of flavoring material is considerably affected b y the presence of various promoting or depressing substances. For example, sucrose has a pronounced effect on the threshold value of sodium chloride. In a mixture of both ingredients, the presence of sucrose at a concentration of below 6% reduces the threshold for salt, while above this value, the sensitivity to salt is diminished (Bujas, 1934). On the other hand, the sweetness of sucrose was pronouncedly depressed by the presence of citric acid at very low concentrations (Pangborn, 1961). Clearly, the flavoring activity of a certain substance may be subjected to the enhancement or depression of other compounds even at very minute concentrations. In practice, a flavor en- hancer may be defined as a seasoning material that improves the flavoring properties of a particular food product, sharpening and emphasizing flavors already present, without adding flavors of their own (Kurtzman and Sjiistriim, 1964; Stier et UZ., 1967).

One of the early products of fermentation to be commercially promoted for the enhancement of flavor, was monosodium glutamate (MSG). This product is now widely used in food technology to emphasize meaty and other flavors. The exact nature of the activity of MSG is still unknown, although several controversial hypotheses have been suggested (Amerine et al., 1965). The picture was further


complicated when the flavor-enhancing activity of certain nucleo- tides was discovered. This may be due to a demasking action on certain receptors, thus exposing more sites to flavor sensation (Beidler, 1966). It seems that not only do these products affect foods by emphasizing present flavors, but also they indirectly influence their taste activity by changing the viscosity of certain foodstuffs (Wagner et al., 1963; Caul and Raymond, 1964).

As pointed out earlier, flavor comprises a number of factors. So far attention has been given mostly to taste. Obviously, odor plays a prominent part in the sensation of flavor. In order that odor be per- ceived, molecules have to be accessible to the olfactory receptors. In other words, a certain degree of volatility of odoring materials is implied.

The olfactory receptors or odor cells are located in the olfactory region in the roof of the nasal cavity. There are about 10 to 20 million cells, from each, minute cilia protrude into the cavity. Axons from the olfactory cells constitute the olfactory nerve that leads to the olfactory bulb of the brain.

There are many theories to explain the mechanism of olfaction; Moncrieff (1951) mentions twenty-four of them. Today, the most commonly held view on the mechanism of olfaction is that of Davies and Taylor (1954), which in essence applies to the induction of the odor stimulus, a mechanism similar to that suggested by Beidler for the perception of taste.

There have been many attempts to classify odors and to establish their basic components. Man can distinguish between an enormous number of different odors. However, Amoore (1952) believes only in seven primary odors: (1) etheral, (2) camphoraceous, (3) musky, (4) floral, (5) pepperminty, (6) pungent, and (7) repulsive. Other odors are considered to be the product of a combination of the primaries. He further believes that the sites of olfaction for different basic odors are distinct (Amoore et at., 1964).

The sensitivity to odors is much more pronounced than that of taste. Some threshold values will be given for comparison (Moncrieff, 1951):

Diethyl ether Vanillin mg./111.~ of air

1 mg./m.3 of air

The different sensitivities of taste and odor can be demonstrated by ethyl alcohol with a threshold for odor of about 4 mg./liter air; while the threshold for taste would approach 130 mg./liter water.

40 P. MARGALITH AND Y . SCIIWAHTZ

Little can be said about the correlation of the chemical structure of odor active compounds and odor perception. Similarly to taste, there are substances of completely different structures that elicit similar odor sensations, while others, with similar chemical constitutions produce completely different odors. The reader is referred to Moncrieff (1951) for a complete discussion of the problems related to chemical constitution and odor. Stereochemical aspects in the physiology of odor perception have been recently suggested by Amoore and Venstrom (1966).

In dealing with flavor it is necessary to define the techniques of flavor evaluation. As previously stated, flavor is primarily the reaction of an individual to a certain compound via his sensory organs. Hence, the evaluation of flavor is greatly influenced by the person exposed to it. Age, sex, and culture as well as mood and alertness, all affect the results of a certain flavor test.

There are a number of tests for the sensory evaluation of flavor, some more suitable for a certain case than others. Generally these tests are carried out by a panel of people employing one or more of the following methods: (1) difference; (2) rank order; (3) scoring; (4) descriptive; (5) acceptance and preference. For a complete evaluation of these sensory tests in the description and evaluation of flavor, the reader is referred to a number of texts (Arthur D. Little, Inc., 1958; Amerine et at., 1865; Moncrieff, 1966). A statistical treatment of results is usually employed in order to reach valid conclusions.

Coming hack to the main topic of this review it seems appropriate to recall the classic paper by Omelianski who as early as 1923 collected information on the “aroma-producing microorganisms” - a designation today commonly used by microbiologists. In this paper Omelianski describes a long list of organisms related to the production of odor, from the sweetish scent that resembles limetree flowers of Pseudo- monas p~ocyanea to the pungent odor of sweat formed by a culture of Bacillusfitzianus. For the sake of piquancy we would like to mention an unusual case, when Omelianski isolated the so-called Bacterium esteroaroma ticum from rabbit brain which, in culture, produced a pleasant fruity aroma resembling the odor of apples!

II. The Contribution of Microorganisms to the Production a n d Development of Flavor in Traditional Foods

A. BAKED PRODUCTS

Bread is probably the most common product of fermentation en-


countered in every day life. The flavor of bread is, however, the result of both the fermentation process in the dough and the crust formation during baking. In addition, the ingredients of the dough contribute to the final flavor of the baked product.

In ancient times and practically up to the beginning of this century, the fermentation process of the dough consisted of a spontaneous fermentation. Part of the successful dough leading to the desirable product was then conserved and used for successive dough fermentation, thus constituting the barm of the sour dough or “sauerteig” of the bakery. Evidently, the fermentation taking place in the dough was due to the activity of a mixed population differing from place to place, and the ingredients employed for any particular bread. Various attempts have been made to analyze the microbial flora in the rising dough. Representatives of the genera Escherichia, Aerobacter, and lactic acid bacteria have been described. A number of organic acids including lactic and acetic acid as well as ethanol, diacetyl, and acetone were found to contribute to the specific flavor of sauerteig (Tanner, 1944). Brewers’ yeast were also employed for dough raising, leading to an excellent bread with a sweet and nutty flavor in the hand of the competent baker (Fance, 1960). When eventually the nature of fermentation became understood and the propagation of bakers’ yeast made their use increasingly popular, the old process of sauerteig was slowly abandoned. Modern bakery equipment as well as obvious sanitary regulations made the exclusive use of bakers’ yeast a necessity. Nevertheless, even today some bakeries employ the sauerteig process for the production of the characteristically flavored rye bread (Fance, 1960).

In discussing the nature of dough raising as a spontaneous process, it is pertinent to mention briefly the problem of leavened and unleavened bread on the occasion of the Passover festivities celebrated by the Hebrews. The Law of Moses forbids the use of leavened bread during Passover. A practical point is however, how to define what may be considered unleavened bread. The Sages of the Babylon Talmud (Pessachim 46a) decided that leavening starts within the time required for walking 1 mile, i.e., approximately 18 minutes after the flour was brought in contact with water. Incidentally, this time evaluation is very close to the generation time of a rapidly dividing bacterium, e.g., Escherichia coli.

Modern breadmaking employs several different processes for the production of bread. Three main types are currently distinguished.

(1) The “Straight-Dough” Method. This involves only one mixing of all ingredients and a fermentation period of 2-4 hours before the

42 P. MARGALI’I’EI AND Y. SCHWARTZ

dough is divided and molded. This method however results in a nonuniform loaf and a less soft product. The quickest of the straight- dough methods would be the “no-time dough,” where no fermentation time is allowed after mixing the ingredients. The dough is immediately molded and processed. This however results in a bread with poor flavor, in spite of the comparatively high concentration of yeast, which contributes to a yeasty taste of the bread but does not improve its flavor. This method is limited for emergency baking only.

(2) The “Sponge and Dough” Method. This involves two distinct operations; first about 60% of the flour and water and the whole load of yeast are mixed and allowed to ferment for 3-5 hours; after that the sponge is mixed with the remaining ingredients. After an addi- tional rest period the dough is molded and processed. Such loaves are usually of better quality in both physicaI appearance and taste.

(3 ) The “Pre-ferment” Method. This is a comparatively recent technique developed for the use in continuous mixing. The pre-ferment consists usually of a mixture of yeasts, sugar, salts or milk, and water. The suspension is thus held for a number of hours, usually 6 at 100”F., with agitation, before it is mixed with the flour (Miller arid Johnson, 1958; Matz, 1960; Skovholt, 1964).

Yeasts fermenting sugars produce ethanol and large amounts of carbon dioxide. The primary role of such a fermentation is, of course, the buildup of the voluminous texture of the loaf due to the form a t’ ion of cells contained by the stretched gluten of the flour. Microbial activity in dough is, however, not limited to the metabolism of yeasts since it is well known that any commercially produced compressed yeast contain large amounts of a variety of bacteria. White (19S4) claims that compressed yeast may tolerate up to lo7 bacteria per gram. This number is likely to increase during dough making as a result of reproduction and infection from other ingredients as well as bakery utensils. Moreover, this accidental bacterial population constitutes an important part of bread flavor. Carlin (1958) has shown that dough raised with a pure culture of bakers’ yeast yielded a loaf of bread with poor taste. However, most investigations dealing with the evolution of flavor during dough raising do not distinguish between the part played by the yeast and that of other microorganisms in the formation of taste and odor.

A great number of different chemical entities were shown to be involved in the constitution of bread flavor. Most of these compounds are organic acids, alcohols, carbonyl compounds, and esters. Wise- blatt (1960) using a gas chromatographic procedure, showed that fermented dough made up in the laboratory, contained the following


acids in decreasing amounts: acetic, butyric, isovaleric, and caproic acids. In an examination of the organic acids in a “straight dough,” in addition to acetic acid, only traces of propionic acid were found (Wiseblatt and Kohn, 1960). In “pre-ferments” a large variety of organic acids were detected comprising, in addition to the afore- mentioned, formic, isobutyric, valeric, crotonic, isocaproic, heptylic, caprylic, pelargonic, capric, lauric, myristic, and palmitic acids (Hunter et al., 1961). Lactic and pyruvic acid were found by other workers (Cole et al., 1962; Johnson et al., 1958). Most of these acids were claimed to be produced during the first hours of the fermentation of the pre-ferment mixture.

Most of the nonacid fractions of dough flavor constituents were studied during recent years in various preparations of pre-ferments. Cole et al. (1962, 1963) studied the kinetics of ethanol production in these systems. n-Propyl, isobutyl, amyl, isoamyl alcohols, as well as 2J-butanedio1, and 2-phenylethyl alcohols were also found in pre- ferment liquids (Smith and C o h a n , 1960). These workers believe, however, that lower alcohols from pre-ferments were not involved in the ultimate flavor of baked bread.

Carbonyl compounds formed during panary fermentation received the attention of many bread chemists. Kohn et al. (1961) investigated the carbonyl compounds from straight dough prepared with compressed yeast and with a special yeast preparation with a very low bacterial count, and made some very important observations with regard to the role of the different groups of organisms in the production of carbonyl compounds during fermentation. The following car- bony1 compounds were analyzed: 2-hexanone7 n-hexanal, isovaleryl- aldehyde, n-butyraldehyde, acetone, and acetaldehyde. A total of about 400 p.p.m. of carbonyl compounds were found in both preparations, with very little difference in the qualitative and quantitative spectrum of the individual carbonyl compounds. It was concluded that bacterial activity contributes little, if at all, to the formation of these compounds, which probably are derived from the activities of the yeast. Miller and co-workers (1961) who studied the carbonyl content of pre-ferments, found in addition to the above compounds also: formaldehyde, isobutyraldehyde, methylethyl ketone, n- valeraldehyde as well as 2-methyl-l-butanal, even though no attempt was made to study the origin of these compounds. Linko et al. (1962) extended this list of carbonyls detected in pre-ferments by adding propionaldehyde and n-hexaldehyde. Other workers identified acetoin (acetylmethylcarbinol) and diacetyl in pre-ferments (Smith and C o h a n , 1960). It seems, however, that the amount of carbonyl com-

44 P. MARGALITH AND Y . SCHWAHTZ

pounds detected in pre-ferments are very inferior to those of ordinary dough. Linko et al. (1962) further examined the effect of various bacterial cultures added to pre-ferments on their carbonyl content. It was found that of all the cultures studied only Pediococcus cere- visiue increased considerably the amount of propionaldehyde and ace tone .

The presence of esters in pre-ferments has been studied by a number of workers. Obviously, ethyl esters predominate. Ethyl formate, ethyl acetate, ethyl lactate and 1,Spropanediol monoacetate have been detected in pre-ferments (Johnson et al., 1958; Smith and CofFman, 1960). However, there is little indication what part, if any, microorganisms take in the formation of these esters.

It should be pointed out that compounds produced by different organisms during the panary fermentation, contribute only in part to the final flavor of bread owing to their great volatility and their disappearance during baking. Other flavor components are produced during crust formation to which microbial activities can contribute only indirectly, e.g., amino acids liberated during proteolysis of dough ingredients, will enhance the browning reactions taking place during baking. Also, Wiseblatt and Zoumut (1963) have shown that the reaction product between proline, abundant in flour protein, and di- hydroxyacetone derived from the panary fermentation, is a compound with a characteristic crackerlike flavor that may constitute a major component of the organoleptic qualities of baked products.

A number of interesting experiments have been carried out by different workers aiming at the flavor improvenient of baked products by the introduction of pure cultures of various bacteria isolated from different sources, which may be involved in the enhancement of bread flavor. Robinson et al. (1958) and Miller and Johnson (1958) examined the effect of pure cultures when added to the pre-ferment. These cultures were isolated from pre-ferments, aged sponge dough, and from dairy sources. Organoleptic examinations of bread, baked with such pre-ferments were carried out. It was found that certain bacteria improved the taste and odor characteristics of bread, especially Lacto- bacillus bulguricus 09, L. pluntarum, and L. breuis. However, no analysis of flavor components of such loaves were recorded. Similarly, Carlin (1959) investigated the flavor fortification of straight and sponge dough by the introduction of various bacteria from the genera Leuco- nostoc and Lactobucillus, previously isolated from commercial yeast, and found a significant improvement of bread flavor. Thus, it can be said that although modern bread making has abandoned the old pro-


cedure of sauerteig in favor of the compressed yeast method, the contribution of the bacterial flora to the development of the characteristic bread flavor has been well documented. The more recent attempts to introduce pure bacterial cultures to the yeast mix may therefore be regarded as a semisynthetic version of the original sauerteig procedure.

B. FERMENTED BEVERAGES

2 . Brewing and Flavor

Modern beer making has led to a change in the production of beer, not only with regard to the amounts of beer brewed in various countries but also in the quality of its taste and aroma; unfortunately, not always for the better. The introduction of the continuous fermentation of beer has so far only aggravated this problem. A better knowledge of the factors involved in the production and enhancement of beer flavor is therefore imperative.

It is evident that the flavor of beer is due to a number of different factors originating in the raw materials used, e.g., the type of malt, nature and amount of adjuncts, hops etc., type of fermentation, aging, and processing. Since we are mainly concerned with the contribution of microorganisms to flavor, the reader is referred to a number of textbooks and more recent reviews for further information on the non- microbial factors involved in the formation of beer flavor (Hind, 1950; DeClerk, 1958).

The primary contribution of brewers’ yeast to the taste of beers is, of course, ethanol. Although the ethanol content varies with the wort employed in the fermentation of the different beer types (ca. 2-6% w./w.) it is more likely that the characteristic flavor of a certain beer is due, to a greater extent, to other products of fermentation. It is generally agreed that higher alcohols (fuse1 oil), aldehydes, ketones, lower fatty acids, and esters, etc., determine the flavor quality of a certain brand. Employing conventional methods of analysis, Hartong (1963) suggested an analytical profile of beer aroma composed of the following groups: (a) higher alcohols, esters, aldehydes, and volatile acids, occurring at mg./liter concentrations, yielding a more or less favorable beverage; (b) diacetyl, hydrogen sulfide, and mercaptans appearing at pg./Iiter quantities that confer an unpleasant aroma to the product. Recent advances in the application of gas-liquid chromatography has greatly advanced our knowledge of the minor components of beer and their relationship to its flavor. Bavisotto and Roch

46 P. MAHCALITH AND Y. SCHWARTZ

(1959) were probably the first to have studied the fermentation of beer employing gas chromatography. Following the formation of fusel oil components it was found that isoamyl alcohol could be identified very soon after the pitching of the wort, reaching maximal values at the time of the major fermentation period, i.e., within 3-5 days, comprising about 60 to 70% of the total fusel oil (Harold et al., 1961; Arkima and Shito, 1963). Pfenninger (1963) analyzed a large number of Swiss beers and found the fusel oil content to vary between 73 and 129 mg./liter. Further, a strong correlation between ethanol production and fusel oil formation could be established, thus confirm- ing the metabolic relationship between these products of yeast activity. A number of interesting points with regard to the nature of beer fermentation and its effect on the formation of fusel oil were raised by Hough and Stevens (1961) who found that top beer always contained higher values of higher alcohols that bottom beer. Further, it was found that fermentations carried out at higher temperatures yielded higher levels of fusel oil. Similar results were obtained by Drews et al. (1964) who examined a large number of German lager beers. A change in temperature of fermentation from 7.5” to 1O.o”C. led to an increase in the higher alcohol content of the final product from 59 to 77 p.p.m. It was further claimed that agitation during beer fermentation increased the level of higher alcohols. Skating and Venema (1961) consider that in rating the characteristic beer flavor, the relative amounts of the alcohols is of importance. Analyzing Dutch beers, the amyl alcohols, i.e., isoamyl and optically active amyl alcohols, were followed in importance by phenethyl alcohol, isobutanol, n-butanol, n-propanol, isopropanol, n-hexanol, and fur- fury1 alcohol.

The occurrence of aromatic alcohols in beer has been studied using gas chromatographic methods by Drews and co-workers (1965), who found in addition to phenethyl alcohol, also tryptophol and tyrosol. The concentration of aromatic alcohols varied in the German beers examined, according to the composition of the wort and temperature of the fermentation. Highest levels of phenethyl alcohols (44 p.p.m.), tryptophol (4 p,p.m.), and tyrosol (28 p.1J.m.) were found in 3 Pilsner beers obtained by an “intensive fermentation.”

It is generally agreed that the biogenesis of higher alcohols during beer fermentation follows the Ehrlich scheme for the transformation of amino acids or their precursors into alcohols via deamination and decarboxylation. The importance of nitrogen constituents of the medium on the regulation of the biosynthesis of higher alcohols via the Ehrlich scheme, has been discussed at length by Ayrapaa (1963).


Relatively little work has been done so far on the effect of yeast species and varieties on the final composition of flavor-determining components of the beer fermentation. Jenard and Devreux (1964) examined a number of strains of Saccharomyces cerevisiae and S. carlsbergensis and found significant variations between the formation of higher alcohols at the stage of the main beer fermentation. Bottom beer brewed with different yeast strains produced total higher alcohols in the range of 59 to 90 p.p.m., whereas top yeast yielded values rang- ing between 101 and 180 p.p.m. Drews et al. (1964) studied the effect of the flocculent nature of certain yeasts on the formation of higher alcohols and found that powdery yeast gave less n-propanol and more 2-methyl-l-butanol than flocculent yeasts.

Ingraham and co-workers (1961) prepared a number of auxotrophic yeast mutants and studied their fermentation metabolites. Significant differences were found between the yields of the different alcohols. A leucineless mutant could not synthesize isoamyl alcohol, while an iso- leucineless strain could not form 2-methylbutanol. A triple auxotroph requiring leucine, isoleucine, and valine would not produce isoamyl, 2-methylbutanol, and isobutanol but formed large amounts of n- butanol, which the wild type produces only in trace amounts. These observations may become of great importance in future studies on the development of better flavor under controlled beer fermentation. The glycerol content of various beers was investigated by Enebo (1957) who found it to vary between 1500 and 2000 mg./liter although little is known on the factors affecting glycerol production during brewing. A number of organic acids comprising acetic, formic, lactic, and traces of other acids, were found in beer. The origin of these compounds has not yet been established although there is good reason to assume that they are derived mainly from the activity of various contaminants during the initial fermentation.

Ethyl acetate is the most abundant ester in beer followed by smaller amounts of ethyl formate, isoamyl acetate, and other esters. It is assumed that yeasts take an active part in the formation of these esters (Harold et al., 1961; Kepner et al., 1963; Masschelein et al., 1965). West et al. (1951) analyzed a large number of American beers and gave an average of 40 p.p.m. for total esters in beer; while Jenard and Devreux (1964) give the maximum of 27 p.p.m. for bottom beer and 82 p.p.m. for top beer. A “fruity” flavor has been described for beers with high ethyl acetate and isoamyl acetate levels (Hartl, 1964; Masschelein et al., 1965; Gilliland and Harrison, 1966).

Acetaldehyde is considered to be a product of leakage of the alcoholic fermentation and attains considerable levels in beers. Maximum

48 P. MARGALITH AND Y. SCHWARTZ

values of acetaldehyde have been recorded during the first days of the main fermentation. The “green” off-flavor of beer has been attributed to high acetaldehyde levels. The disappearance of this off-flavor during the latter part of the fermentation and storage has been correlated to a decrease in acetaldehyde (Enebo, 1957; Bavisotto and Roch, 1960; Sandegren and Enebo, 1961). Although diacetyl occurs in beer only at very low concentrations, it may be the cause of various off-flavors. West and co-workers (1952) analyzed a large number of beers with normal and off-flavor. They found that diacetyl at concentrations above 0.5 p.p.m. is responsible for these off-flavors. Although diacetyl is considered to be a normal product of yeast alcoholic fermentation, the excessive formation of diacetyl leading to the well known off-flavors are probably due to the activity of contaminants such as Pediococcus, Aerobacter, etc. (Sandegren and Enebo, 1961). However, Czarnecki and Van Engel (1959), and Gilliland and Harrison (1966) pointed out that high values for diacetyl in beer could be obtained also when respiratory-deficient mutants of S . cerevisiae were used. Gjertsen e t nl. (1964) believe that more diacetyl is produced during the fermentation with flocculent yeast than with powdery strains. Brenner et al. (1963) claim that the nutritional composition of the wort has a significant effect on the diacetyl formation during fermentation. A number of methods have been suggested to reduce the diacetyl content of beers in order to reach the organoleptically acceptable levels of below 0.2-0.3 p.p.m. From the present standpoint the most interesting method would be the practice of “Krausening” which leads to a more reductive environment (Sandegren and Enebo, 1961; Lawrence, 1964)-

The biogenesis of diacetyl in beer has been studied by a number of workers. West et al. (1952) considered dimethylene glycol to be the precursor of acetoin (acetylmethylcarbinol), and diacetyl. However, Sandegren and Enebo (1961) in a later report, suggested that acetoin is derived from the condensation of 2 moles of acetaldehyde; acetoin being oxidized to form diacetyl. A similar mechanism was suggested by Antoniani (1961). On the other hand, Owades et aZ. (1959) and Yoshizawa (1964) suggest a scheme in which alanine serves as initial source for the formation of pyruvate which yields a-acetolactic acid and acetoin. Aerobic conditions favor the transformation of acetoin into diacetyl (Owades et al., 1959).

In practice, acetoin in European beers was found within the range of 3.1 and 14.6 p.p.m. A number of factors seem to affect the formation of acetoin. Lower vitamin levels in wort as well as a deficiency in Mg ions seem to favor acetoin formation (Yoshizawa, 1964). An inositol-

FLAVOR AND MZCROORGANISMS 49

deficient strain was found to produce much higher concentrations of this compound (Lewin and Smith, 1984).

Sulfur-containing volatiles are usually present in most beers at very low concentrations (p.p.b.) and are generally considered to be produced during the active part of the yeast fermentation. These compounds comprise hydrogen sulfide, mercaptans, thioformaldehyde, dithioformaldehyde, and thioacetone (Hashimoto and Kuroiwa, 1966).

An interesting feature of yeast activity and its detrimental effect on beer flavor has been recently reported. Among the phenolic compounds, ferulic acid seems to be very common in grain mash and final beer. Certain yeasts seem to be involved in the decarboxylation of ferulic acid and the formation of 4-methylguaiacol and 4-vinylguaiacol which gave a pungent off-flavor to the beverage (Brumsted et al., 1965):

(4-Methylguaiacol) OH

(Ferulic acid)

Recent advances in brewing technology and the new practice of continuous fermentation have led a number of workers to study the effect of batch versus continuous processes on the formation of taste and aroma of beers. Hudson and Stevens (1960) found that contin- uously fermented top beers did not differ significantly from those produced by conventional methods. However, in the case of the more common bottom fermentation Sandegren and Enebo (1961) found a number of interesting differences between the two kinds of fermentation. Beer produced by the continuous process contained higher nonvolatile acids and was of lower pH, showed greater accumulation of acetaldehyde, a higher fatty acid content, more higher alcohols,


comprising higher values of phenylethyl alcohol, and apparently also a higher potential for the formation of sulfhydryl compounds. Nord- strom (1965) also found an increased production of volatile esters in the continuous process. In general, it is considered that the continuous fermentation of bottom beer bestows a more alelike flavor to the final product (Sandegren and Enebo, 1961).

An interesting case of flavor formation in beer is the use of Brettano- myces and Lactobacillus pastorianus in the secondary fermentation for the production of the Lambic-type beer. The formation of the characteristic flavor is mainly due to the increase in acidity and was found to be optimal with a wort of comparatively high specific gravity (Gilliland, 1961). However, little is known on other flavoring compounds formed during such a secondary fermentation.

In summary it may be said that microorganisms determine to a great extent the characteristic flavor of the final product. Recent advances in flavor analysis have opened new fields for the characterization of various beer types, although no definite correlations have so far been laid down.

2. Vinijication and Flavor

Making wine is a very old practice. To cover all the technological developments which led to modern wine making is outside the scope of the present review. We shall limit ourselves to microbial activities concerned with wine making and their contribution to flavor. There are many textbooks and reviews which deal with this problem at vary- ing lengths so that much of the material will be only briefly sum- marized or referred to.

Similarly to brewing the taste and aroma of a bottle of wine is derived from a number of factors affecting the final product: raw materials, technology of vinification, and processing. However, wine making differs from brewing not only in raw materials and fermentation technology, but also in another aspect, i.e., in the less rigid control and hence much greater variability in all the stages of wine making. Climatic differences and a variety of agrotechnical factors have a pronounced effect on the nature of the chemical composition of must. It is, therefore, not surprising that many workers believe that the major contribution to flavor is due to the nature of the grape variety and climatic conditions of the corresponding vintage and only to a lesser part to the character of the microorganism involved during vinification. However, in recent years and with the advent of modern analytical approaches, the importance of the microorganism both dur-


ing the main vinification and the secondary fermentation processes have been newly emphasized.

The main product of vinification is of course ethanol. Although wine is characterized primarily by its alcohol content (that varies according to the sugar content of the must and fermentation technology, but may reach much higher concentrations than in beer and other alcoholic beverages), it is the emphasis on diminution of other bouquet components that determine the importance of ethanol to the perception of flavor. Thus, it has been recently shown by Rankine (1967) that the threshold values of a large number of higher alcohols, the importance of which is well known with regard to flavor evaluation, is increased over a hundredfold as compared to the values obtained in distilled water for the corresponding alcohols.

Most of the higher aliphatic alcohols discerned in beer fermentation take part also in the higher alcohol composition of wine. The most important.ingredients of the fusel oil are: is0 and active amyl alcohols, isobutanol, and n-propanol. The most objectionable, with regard to taste and odor, is that of isoamyl alcohol. The average fusel oil content of wines has been examined by a large number of workers. In general the higher alcohol content of red wines is higher than that of white wines. Guymon and Heitz (1952) give a mean value of 250 p.p.m. for white Californian table wines versus 287 p.p.m. for red table wines, while Peynaud and Guimberteau (1962) find 309 p.p.m. and 394 p.p.m. (mean values) for the corresponding types of a large number of French wines. Although it is generally agreed that the fermenting yeast is responsible for the formation of the higher alcohols (Guymon, 1966), until the last decade comparatively little was done to examine the fusel oil-producing capacity of various yeast strains involved in vinification. However, as pointed out by Webb (1967) “the group of alcohols is usually present at concentrations low enough so that the sensory impression is not unfavorable.”

The capacity of various yeasts to produce fusel oil components has been examined by a number of workers. Webb and Ingraham (1963) in their exhaustive review on the problem of fusel oil, summarize earlier observations and state that although various yeast genera comprising a number of wild yeasts, produce different amounts of fusel oil, little variation in the formation of fusel oil could be found in the naturally occurring wine yeasts. Nevertheless, Webb and Kepner (1961) noted considerable differences in the composition of fusel oil of wines that were fermented with three different wine yeasts with the same grape juice and under identical conditions of vinification.


The highest level of n-propanol was found in the wine fermented with a Jerez yeast (20.2 weight % of total fuse1 oil) as compared to 2.6% of wine fermented with Montrachet yeast. Amy1 alcohols were highest in Montrachet fermented wine (94.7%) and lowest with Bur- gundy yeast (69.4%); Montrachet yeast produced the largest amount of isobutyl alcohol (16.5%). The most recent analysis of the ability of various species of Saccharomyces to produce higher alcohols has been reported by Rankine (1967). Vinification was performed in a number of grape juices employing different yeasts and the formation of alcohols was analyzed for each species. Significant differences were found between the different organisms. The following ranges were observed. (Means from different grape juices with same yeast.)

n-propyl alcohol: minimum 13 p.p.m, (S. carlsbegensis No. 731) maximum 106 p.p.m. ( S . cereuisiae No. 350)

isobutyl alcohol: minimum 9 p.p.m. (S. fructuum No. 138) maximum 34 p p m . (S. cheunlieri No. 317)

amyl alcohols: minimum 115 p.p.m. ( S . cereuisiae No. 213) maximum 262 p.p.m. (S. cereuisiae No. 727)

The production by the various yeasts relative to one another seems to be consi5tent so that it would be justified to designate certain yeasts as high or low producer of some of the higher alcohols. Furthermore, it may be concluded that different strains of the same species seem to differ substantially in their capability to produce these alcohols. Guymon (1966) has also shown that mutants of S. cereuisiae would produce different amounts of the various higher alcohols according to their blocks in the biosynthetic pathways of their corresponding precursors. This seems to be in perfect agreement with results obtained in a similar approach with beer-fermenting yeasts (Ingraham et al., 1961). The reader is referred to Webb and Ingraham (1963) and Guymon (1966) for further information on our present knowledge on the biogenesis of various higher alcohols in wine fermentation.

It is evident that organic acids have a very important place in the formation of aroma of wine. Recent analyses employing gas-liquid chromatography have revealed a large number of acids during vinification and maturation. Van Wyk and co-workers (1967) have published the acid content of a methylene chloride extract of White Riesling and found: acetic, n-butyric, n-caproic, n-caprylic, n-capric, 9-decenoic, and succinic acids as major constituents. Formic, propionic, isobutyric, 2-methylbutyric, isovaleric, lactic, 2-hydroxyisocaproic, n-pelargonic, and malic acids were found in smaller amounts. However, no dis- tinction between the origin of these acids, if derived from the grape


juice or formed during vinification, has been suggested. Volatile acids, primarily acetic acid, are produced during bacterial-free alcoholic fermentation in considerable amounts. Amerine and Cruess (1960) give the values of 0.03-0.05% in different wines as compared to 0-0.02% in the respective musts. The excess of acetic acid, over 0.05% is an indication of bacterial spoilage. However, a number of yeasts are known to possess a greater volatile acid-producing ability. Cappucci (1948) claimed to have found up to 0.369% volatile acidity in normal fresh wines and attributed this to the activity of apiculated yeasts and certain Zygosaccharomyces. Although wine spoilage is generally considered to be due to the formation of excess acetic acid, it has been shown experimentally that the organoleptic deterioration of wine is not due to the accumulation of volatile acids, but rather to the formation of high levels of ethyl acetate (Amerine, 1954; Ribereau-Gayon and Peynaud, 1961).

With regard to the biogenesis of acetic acid during the yeast fermentation, it seems that, oxidation-reduction potentials are involved in the oxidation of acetaldehyde to acetic acid and its reduction. It has been pointed out that during the initial stages of fermentation, levels of acetic acid are considerably higher than toward the final stages (Joslyn and Dunn, 1941; Ribereau-Gayon and Peynaud, 1946). However, no definite evidence has been put forward for the biochemical pathways and enzymes that perform these reactions.

Other volatile acids, such as formic, propionic, and butyric acids, are probably not derived from the yeast fermentation. In spite of the pungent odor of these compounds their occurrence of, only in very minute amounts, in normal wine may contribute favorably to the taste of this beverage (Webb, 1967).

Among the nonvolatile organic acids produced during vinification, lactic acid occupies a very important position, both from the practical standpoint and biochemical interest. Lactic acid occurs in all wines that have undergone the vinification process. Wines that have had no bacterial contamination, contain usually only very low levels, up to 0.06% of lactic acid, which has been recently shown to be predominantly of the D(-) isomer (Peynaud et al., 1966). In a later work (Peynaud and co-workers, 1967) the formation of lactic acid by various yeasts has been examined under controlled conditions of vinification. It was found that there exists considerable variation among the species of Saccharomyces. Saccharomyces rosei gave the lowest level 101-135 mg./liter, while most of the saccharomycetes produced around 200- 400 mg./liter. An interesting exception was found with S. veronae which produced up to 1800 mg./liter and was further distinguished by


the fact that most of the acid formed was of the L(+) isomer. The contribution of lactic acids formed during yeast vinification to the flavor of wine, has so far received little attention.

The most important feature of lactic acid in wines, with all its organoleptic implications is, however, not due to the activity of yeasts but to the formation of lactic acid by certain groups of bacteria. Since the early years of wine microbiology (see Ribkreau-Gayon and Pey- naud, 1961, for an excellent historical review) it was known that certain bacterial processes lead to a change in the acidity of wines after the alcoholic fermentation was completed. A reduction of acidity is of great importance in many cases of table wines that are produced from juices particularly rich in total acidity. According to Amerine and Cruess (1960) almost half of the acidity in various musts is due to the presence of L-malic acid (0.1-0.8%). Malic acid, a relatively strong dicarboxylic acid is readily attacked by many lactobacilli, which thus reduce the acidity of the wine by the production of the weak, mono- carboxylic lactic acid according to:

5 7% NADH, , lactic dehydrase + & CH(OH)COOH

Most of the lactic acid thus produced has been shown to be of the L(+) isomer (Peynaud et d., 1966). Various workers have studied the biochemical mechanisms involved in the decarboxylation of malic acid and the reduction of the intermediate pyruvic acid to form the lactic acid. A number of coenzymes and cofactors have been described (Ochoa, 1951; Jerchel et al., 1956). An important point is sometimes overlooked. Thermodynamically, the reaction that leads to the decrease in malic acid is an energy consuming process, hence the malo- lactic fermentation requires the presence of a source of energy (Schanderl, 1943). However, as pointed out by Radler (1958) the amounts of carbon compounds that drive this reaction are minute and affect little, if at all, the composition and flavor qualities of the wine. This is in sharp contrast to the reduction of malic acid, which may totally disappear from the wine (Pilone and Kunkee, 1965).

Although the involvement of lactobacilli in the malo-lactic fermen-


tation has been known for a long time, the exact nature of the microorganism was studied little until the early fifties. Ribereau-Gayon and Peyiiaud (1961) describe in detail most of the work done until 1960. Hornofernientative and heterofermentative Lactobacillus (L. plantarum, L . delbriikii, L. brevis, and L. hilgardii) as well as homofermentative Pediococcus, ( P . cerevisiae) and heterofermentative Leuco- nostoc (L. citrovorum) have been described in wines that had undergone the malo-lactic fermentation (Lambion and Meskhi, 1957; Ingraham et al., 1960; Fornachon and Lloyd, 1965).

Although the malo-lactic fermentation can be easily demonstrated under laboratory conditions, it occurs frequently but not always under commercial conditions. This is probably due to the inability of the lactic culture to become established owing to a number of factors such as high sulfitage, high alcohol content, and early separation of wine yeasts by racking which thus deprives the wine from the growth factors required by the lactobacilli, etc. (Fornachon, 1957). This author also stressed the importance of the p H as observed with Aus- tralian wines. It seems that not only is the development of the malo- lactic culture sensitive to low pH values (lower limit 2.9-3.0), but the nature of the lactic culture that becomes established differs with the pH. Other factors which control the outcome of the malo-lactic fermentation have been recently reviewed by Kunkee (1967a,b). The time interval between the alcoholic fermentation and the onset of the malo-lactic activity was found to be about 3 weeks (Rice, 1965).

Frequent failures in the commercial application of the malo-lactic fermentation have led many workers to attempt the reduction of acidity in wine by introducing an abundant starter population of a selected lactic organism. French authors, who strongly advocate the use of the malo-lactic induction by massive inocula, expect its large- scale application to be realized in the very near future. Ribereau- Gayon and Peynaud (1961) suggest the use of homofermentative lactobacilli for this purpose. This is in accordance with the widely accepted view that heterofermentatives produce considerable quantities of flavor affecting volatiles, such as acetoin and especially diacetyl, which may become deleterious to wine bouquet (Vaughn and Tchelistcheff, 1957). However, American enologists ?re not in complete agreement with these observations. Pilone et al. (1966), who made a careful study of experimental vinification of Californian wines employing pure lactic cultures and modern procedures of analysis, found that malo-lactic reduction of acidity is carried out by various lactic organisms, with a slight increase in volatile acidity, and changes

56 P. MARCALITH AND Y. SCHWARTZ

in the acetoin + diacetyl values (control 7.5-8.4 p.p.m. versus 4.9- 13.9 p.p.m. in the malo-lactic wines) which do not support the con- clusion that homofermentatives are less productive with regard to diacetyl and acetoin. A number of other compounds seem also to be affected by the malo-lactic fermentation. Significant higher values have been found for ethyl lactate in the malo-lactic wines, while hexyl alcohol appears to diminish. The importance of these changes with regard to flavor has not yet been elucidated. However, as pointed out earlier (Pilone and Kunkee, 1965), while organoleptic tests revealed significant differences between normal and malo-lactic wines, differences between wines produced with a number of different lactic cultures could be distinguished only by some of the members of the taste panel.

Considerable quantities of succinic acid are known to be formed during the alcoholic fermentation. Levels of up to 0.15% have been recorded in wines (Amerine and Cruess, 1960). Little is known about their specific contribution to the taste of wine. Recently some information became available regarding the contribution of various microorganism to the formation of ketonic acids in fermented juice. Lafon- Lafourcade and Peynaud (1966) found up to 785 mg./liter of pyruvic acid and up to 146 mg./liter of a-ketoglutaric acid in wines of different vintages. Yeasts related to the vinification process seem to have different abilities with regard to the production of these keto acids. Low values were obtained with wild yeasts, such as KZoeckera apdcu- Zata, Hansenula anomda (less than 100 and 30 p.p.m., respectively) while strong alcohol-producing yeasts formed higher levels; in the case of Schixosaccharomyces pombe up to 830 and 284 p.p.m., respectively. The effect of these acids on the formation of wine bouquet has not yet been studied.

The importance of esters formed during vinification to the formation of wine aroma has been studied over many years. Ethyl acetate is the most prevalent ester in fermented must and, as pointed out by Amerine (1954), constitutes the only ester important from the standpoint of taste and aroma. Other esters such as ethyl caproate, heptanoate, caprylate, pelargonate, isoamyl acetate, and others have been identified in various wines. Van Wyk et al. (1967) enumerate a total of 22 esters found in Riesling wine.

The content of ethyl acetate in wine varies according to the grapes employed and the vinification process. Ribkreau-Gayon and Peynaud (1961) who analyzed the composition of wines during several decades found wines to contain generally between 44 and 176 mg./liter ethyl


acetate; seemingly, wines can tolerate up to 200 mg./liter of ethyl acetate, and above this value appears to give a spoiled character to the wine (Amerine and Cruess, 1960). It is now generally agreed upon that the formation of esters is due to the metabolic activities of microorganisms and little, if at all, to chemical esterification. Ribbreau- Gayon and Peynaud (1961) distinguish between 5 different groups of microorganisms with regard to their production of ethyl acetate. First, the true wine yeasts of the genus Saccharomyces which never produce more than 50 mg./liter. Second, yeasts from the group of: Candida pulcherima, Kloeckera africana, and Brettanomyces which may enhance the ethyl acetate content of wine up to 110 mg./liter, however, these are of little practical importance. Third, .Saccharomycodes ludwigii which differs from the other yeasts in the fact that its esterogenic activity is greatly enhanced under anaerobic conditions. Grape juice fermented with S . ludwigii has a very pronounced odor of this ester (up to 200 mg./liter). Fourth, the abundant group of apiculated wild yeasts of the Kloeckera-Hanseniaspora type which produce over 300 mg./liter of the ester and fifth, the group that comprises yeasts such as Pichia spp. and more important, Hansenula anomala which apparently has the strongest esterogenic activity. The latter was found to produce up to 900 mg./liter of ethyl acetate and thus may be detrimental to the process of vinification. With the exception of Saccharomyeodes ludwigii all esterogenic activity is promoted by contact with air. Under these conditions various species of Aceto- bacter may proliferate with a concomitant increase in the formation of ethyl acetate. Tabachnick and Joslyn (1953) have studied the kinetics of ethyl acetate formation and found that levels of the ester decrease with time, under aerobic conditions. This seems to be in accordance with the observation of Wahab et al. (1949) who found, that on aging, the ester content of wines decreases substantially with a constant improvement of wine bouquet. Ough and Amerine (1967) report the optimal temperature for ester formation to be in the vicinity of 20°C.

Since early days of wine microbiology it was known that glycerol is a constant by-product of the alcoholic fermentation. Its sweet taste and viscous character undoubtedly has a pronounced effect on the character of the fermented product, especially what the enologist calls “body.” Others believe that glycerol imparts smoothness and amelio- rates the burning taste of alcohol (Hickinbotham and Ryan, 1948; Hinreiner et al., 1955). Glycerol in wines may be detected organoleptically at 0.9% in white wine and 1.3% in red wine (Hinreiner

58 P. MARCALITH AND Y . SCIIWAHTZ

et al., 1955). The usual concentration of glycerol in wines, as given by Amerine (1954), is in the range of 0.5-1.5% and therefore may be regarded as a significant constituent of wine flavor. The effect of temperature of vinification, the concentration of sulfite, sugar, and other factors have been studied by a number of workers (Venezia and Centilini, 1941; Hickinbotham and Ryan, 1948; Uchimoto and Cruess, 1952; Gentilini and Cappelleri, 1959). In the case of botrytized grapes, unfermented juices may contain already considerable quantities of glycerol due to the metabolic activity of the fungus Botrytis cinerea. (For a summary of the information on B. cinerea see Ribereau-Gayon and Peynaud, 1964.)

A number of workers have studied the glycerol-forming ability of various yeast species and strains. Interestingly, strong fermenting yeasts show considerable similarity in their glycerol production under experimental conditions. Usually levels of 0.6-0.9% are found. Zygosaccharomyces acidifaciens seems to be outstanding with respect to its glycerol-prod\icing capacity, since it was found to yield over 1.5% under similar conditions (Ribereau-Gayon and Peynaud, 1964).

Do iiorrnal wine yeasts contribute to the level of carbonyl compounds during vinification? There seems to be little agreement with regard to the quantities of acetaldehyde found in wines. Ribkreau-Gayon and Peynaud (1961) give the range of 40-120 p.p.m., while Ough and Amerine (1967) found much lower values, 6-49 p.p.m. These differences may be due to different analytical procedures, but also may de- pend on the concentration of sulfite during vinification. Sulfite, that entraps acetaldehyde, prevents its reduction and thus leads to higher levels. However, as pointed out by Hinreiner et al. (1955), at these concentrations acetaldehyde in wine contributes little to the organoleptic qualities of the beverage, the threshold levels being 100-125 p.p.m. in red and white table wines, respectively. Acetoin and diacetyl are found at much lower concentrations. Fornachon and Lloyd (1965) give the values of 0.7-0.9 pap.m. and 0.1-0.3 p.p.m., respectively, for table wines, as against the much higher levels found by Ribereau-Gayon and Peynaud (1961) 4-25 p.p.m. and 0.5-2.5 p.p.m., respectively. Clearly, various factors such as temperature of vinification, aeration, and agitation strongly affect the levels of these carbonyl compounds (Crowell and Guymon, 1963; Ough and Amerine, 1967). Carefully controlled experiments conducted by Guymon and Crowell (1965) have shown that both sugar solutions and grape juice inoculated with a pure culture of S. cerevisiae (Montrachet strain) produced acetoin and diacetyl, reaching a maximum of 25-100 p.p.m.


(combined) about midway through the vinification process, while declining strongly toward the end of the fermentation. As expected, fortified wines yielded much higher values (up to 236 p.p.m.). From the organoleptic standpoint, both acetoin and its reduction product 2,3-butylene glycol, have apparently little importance, although the former is considered to be the precursor of its oxidized form diacetyl, which may be detected at very low levels. According to Amerine (1954) diacetyl in wines may be detected already at 2-4 p.p.m. levels, which as pointed out earlier, may be reached in certain normal wine fermentations. Whether very low diacetyl levels formed during vinification may contribute favorably to the characteristic qualities of certain wines is still a matter of controversy.

Hydrogen sulfide and mercaptans in fresh wines may be frequently encountered (Amerine and Cruess, 1960). Considering the very low threshold values of these compounds [0.1-1.0 p.p.m. for hydrogen sulfide, according to Staudenmayer (1961)], their occurrence in wine are of primary importance in the sensory evaluation of this beverage. Various workers have attempted to reveal the origin and mode of formation of these compounds during vinification. It is generally assumed that most of the hydrogen sulfide is derived from sulfur occurring on sprayed grapes or used in the wineries, and to much lesser extent due to the reduction of sulfur dioxide or bisulfite (Rent- schler, 1951). However, other workers also consider organic compounds such as sulfur-containing amino acids, metabolized during vinification, to contribute significantly to the accumulation of reduced sulfur compounds (Benvegnin et d., 1951; Woll, 1955). From our standpoint, however, it would be interesting to elucidate the contribution of yeasts employed during the vinification process, to the formation and accumulation of such undesirable by-products of fermentation. Much of the information available today on this problem is due to the thorough investigation carried out by Rankine (1963). In a number of carefully controlled experiments followed by gas-liquid chromatographic analysis, it was found that the accumulation of H2S during fermentation is strongly correlated to the decrease in the oxidation-reduction potential during the alcoholic fermentation. The yeast varieties employed in these experiments displayed considerable variation with regard to the HzS-forming ability. Values of 0.0 to 7.0 p.p.m., in the absence of any inorganic sulfur compounds, and up to 12 p.p.m. in the presence of elemental sulfur, were obtained. Rankine concludes that most of the hydrogen sulfide in wines is derived from the reduction of elemental sulfur and only to a lesser extent to other inorganic and organic compounds. Ethanethiol, which is probably the


most deleterious compound, is considered to be formed from H2S and ethanol, although the mechanism of its production has still to be worked out. In spite of the occurrence of different yeasts with different hydrogen sulfide-forming abilities, and the practical aspects which may be derived from this fact, it is not known whether under commercial conditions of vinification, yeast are directly responsible for the sulfur reductive reaction or only indirectly through the low E h values developed during fermentation.

One of the most interesting cases in which a microbial process leads to the transformation of a fermented product into a beverage of higher organoleptic and commercial qualities, is the flor type sherry. This wine originated in the region of Jerez de la Frontera where the normal alcoholic fermentation (14.5-15.5% volume alcohol-for the characteristic fino-type) is followed by a secondary process carried out in the so-called solera-system. There, in a number of barrels filled only about 80% of their volume, the wine undergoes an oxidative process owing to the development of a certain film yeast-theflor yeast. Fresh wine is introduced into such a system, transferred from one layer of barrels to a successive one, about six times, until the characteristic flavor of the flor sherry is acquired. For a detailed description of the process see Bobadilla (1943) and Schanderl (1959). Recent advances in wine technology have introduced other methods of sherry production with the aim of eliminating the time consuming solera procedure, by either promoting the flor yeast under submerged culture condition with occasional agitation and aeration or by the “baked process” without the intervention of the microbial stage (Ough and Amerine, 1960).

There is little doubt, that this very complicated wine with its distinct aroma, as produced by the traditional Spanish solera system, is a result of the activity of the typical flor yeast. There exists some argu- ment with regard to the speciation and nomenclature of this yeast. Some workers believe flor yeast to belong to a special species: Sac- charomyces beticus (Marcilla et al., 1936) or s. fermentati (Rankine, 1955; Castor and Archer, 1957) while others believe that the flor yeast is but the oxidative stage of a normal wine yeast, of which many strains form the pellicle stage when cultured under suitable conditions (Schanderl, 1959).

The chemical transformations that the sherry wine undergoes during the oxidative stage are very complex and have been studied by a large number of enologists. The principal flavor components characteristic of sherry wine are aldehydes and acetals. Acetaldehyde is quantitatively the most important. During “sherryzation,” amounts of


over 300 p.p.m. aldehydes may be formed during commercial production (Marcilla et al., 19361, while Amerine (1958) reported much higher levels of about 1000 p.p.m. under experimental condition employing the Californian method of submerged culture. Considering the fact that normal wines contain only minor amounts of aldehydes, it is clear that these compounds contribute significally to the formation of the sherry wine aroma. However, there is still a scarcity of information with regard to the optimal concentration of aldehydes in sherry for the consumers’ market. A blending procedure for the production of sherry with conventional concentrations of acetaldehyde has been proposed by Webb et al. (1964) for the Californian types of submerged sherry.

Acetals are the product of condensation of aldehyde with two molecules of alcohols and the elimination of water:

OCzH, /

\ CHSCHO + 2C2H5OH CHsCH + HzO

OCzHs (Acetaldehyde) (Ethanol) (Diethyl acetal)

Although there are indications of the occurrence of acetals in wine grapes and normal wines (Kepner and Webb, 1956; Webb and Kepner, 1957; Lipis and Mamakova, 1963) few quantitative data are available. However, it is assumed that the “sherryzation” process increases significantly the amounts of acetals present in sherries (Webb et al., 1964), and thus, contribute considerably to the specific character of these wines. With the advent of gas chromatographic analysis further insight has been gained with regard to the qualitative constitution of the acetal compounds. Diethyl acetal was found in higher amounts in flor sherry than in the submerged culture. Galetto et al. (1966) detected eight other acetals: ethyl-active amyl, ethyl-isoamyl, ethylpentyl, isoamyl-pentyl, diactive amyl, active amyl-isoamyl, active amyl-pentyl, and diisoamyl acetal. Although it has been proposed to evaluate the organoleptic qualities of sherries according to their acetal constituents, there is little evidence that the acetals are formed during the microbial stage, since similar amounts of diethyl acetal could be demonstrated in “baked” sherries (Webb et al., 1964).

Both flor sherry and the “baked” product have been shown to contain acetoin in the usual range of 10-32 p.p.m. (Lukas, cited in Schanderl, 1959; Guymon and Crowell, 1965). However, the submerged product is outstanding in its acetoin content which may be as high as 350 p.p.m. Undoubtedly, the aerated culture leads to a higher


metabolic activity of the flor yeast as compared to that of the solera system.

The effect of flor yeast on the accumulation and degradation of alcohols during “sherryzation” has been studied in various systems. Ethanol was found in some cases to decrease somewhat (Schanderl, 1959; Webb et al., 1966) which, at least in part, may be attributed to evaporation. Glycerol has been found to decrease considerably. Marcilla et ul. (1936) claim a 50% reduction of the glycerol content after the completion of the flor process. Similar results were obtained in the wbmerged culture sherry (Amerine, 1958).

In addition to the customary fuse1 compounds, benzyl alcohol, and 2-phenylethyl alcohol were found in flor sherry of Spanish origin (Webb et al., 1967). Numerous publications deal with the occurrence of other alcohols, acids, and esters as determined by gas-liquid chromatography. Lactones, ethyl-.l-hydroxybutyrate, ethyl pyroglutnmate, diethyl siiccinate, diethyl nialate, and 2-phenylethyl caproate were identified. It is suspected that the presence of acetate and caproate esters of 2-phenylethyl alcohol have a special place in the formation of the typical sherry flavor (Webb and Kepner, 1962; Webb et al., 1964; Rodopulo et ul., 1967). However, information on the quantitative arid organoleptic aspects of the sherry aroma components is still fragmentary. It should be pointed out that the contribution of flor yeast to the totii’l aroma edifice of shcrry can be assessed only if uoni- parative analyses are carried out with fresh wine, before and after the termination of the sherry process.

Comparatively little has been published on the production of flavor in nongrape alcoholic beverages. Apple cider is the best known example of such fruit juice that undergoes an alcoholic fermentation (54% v./v., ethanol). Pollard et u1. (1966) reviewed the flavor components of apple cider and pointed out that the most important change during the processing is the transformation of the fruity acid character of the juice to the softer mature flavor of the cider. It has been shown earlier (Whiting and Coggins, 1960) that these chaliges are partly due to bacterial activities during or after the yeast fermentation. Lactic acid bacteria may attack sugars in nonsulfited juices with pH above 3.8. Concurrently, malic acid is converted to lactic acid in a manner similar to the malo-lactic fermentation in wine. Also the yeast fermentation contributes to the general increase in cider acidity due to the significant formation of succinic acid. (This is obscured in the grape juice fermentation owing to the changes in the tartaric acidity in the C02-saturated environment.) There was little difference between the yeast strains employed (Thoukis et ul., 1965).


Another component of microbial origin associated with the desirable flavor of cider is the relatively high level of fusel oil. Although some of these higher alcohols may be present in certain varieties of apple juices, the bulk of this fraction is derived from the yeast fermentation. In comparing the content of higher alcohols of apple juice and cider, it was found that with the exception of n-butanol, alcohols increase significantly during the fermentation, primarily isobutyl, is0 and active amyl alcohols as well as 2-phenylethanol. In a series of laboratory experiments it was shown that the formation of apple juice by its natural microflora usually resulted in a much higher fusel oil content than when the fermentation was carried out with a culture yeast used in this process: 197-335 p.p.m. versus 151-167 p.p.m., respectively; in the case of 2-phenylethanol 127-254 p.p.m. as against 33-65 p.p.m. with the culture yeast. Usually the fusel oil content of ciders are intermediate between that of low gravity beers and wines, which would correlate with their respective ethanol content. However, in the case of cider the scented character of some of these alcohols may constitute an important part of its aroma. Taste panels have shown that up to 200 p.p.m. of higher alcohols were considered essential to typical cider aroma (Pollard et al., 1966).

Sake is the most widely known fermented beverage of the Far East. Production involves a mixed fermentation of rice, employing Asper- gi2lus oryzae for the transformation of the starch into fermentable sugars, as well as a yeast for the alcoholic fermentation, and some bacteria, such as Lactobacillus saki. The characteristic Sake flavor arises thus directly or indirectly from the constituents of rice. Recent analytical studies have shown that most of the fermentation by- products detected in the conventional alcoholic fermentations such as fusel oil, acids, and carbonyl and phenolic compounds, have also been detected in rice wine (Yamamoto, 1961; Komoda et al., 1966; Owaki, 1967). The analysis of Sakk flavor, however, reveals little of the changes that occur due to the chemical activities of each organism involved in the mixed fermentation. A significant increase in the content of tricarboxylic acids as a result of fungal metabolism has been demonstrated during the initial stages of the fermentation (SakC-Koji).

Recently more information has become available on the formation of phenolic compounds during the Sake fermentation. Yamamoto et al. (1961) identified ferulic acid, vanillin, and vanillinic acid in SakC, while in the Sakk-Koji only ferulic acid could be demonstrated. It was assumed, therefore, that the transformation of ferulic acid, which constitutes a minor component of the plant material, was carried out by the microorganisms involved in the later stages of the SakC fermen-


tation. In a later work, Omori and co-workers (1968) describe the metabolism of ferulic acid by a pure culture of a Sake yeast under laboratory conditions. It was found that both p-hydroxybenzoic acid and vanillic acid were formed. When the substrate was changed into vanillin, p-hydroxybenzoic acid, p-hydroxybenzaldehyde, and vanillic acid could be identified. It was concluded that vanillin may be formed as an intermediate in the degradation of ferulic acid, followed by the demethoxylation of vanillin. p-Hydroxybenzoic aldehyde and acid have also been shown to be formed from amino acids like tyrosine by a number of yeasts. However, it is known that vanillin is a more important component of Sake flavor, contributing to its light and sweet fragrance (Yamamoto et al., 1966).

C. DAIRY PRODUCTS

The contribution of microorganisms to the development of specific flavors associated with milk products is widely known, although de- $,tailed studies to provide deeper insight into the chemical entities involved in many processes are still needed.

Milk products such as sour milk, yogurt, cream, and cream cheeses as well as butter, which undergo practically no aging processes, have taste characteristics due in part to the lactic acid produced during the manufacturing. The typical dairy aroma of these products is developed by a specific aroma-producing microflora, predominantly from the heterofermentative group of Leuconostoc organisms. M’ilk citrate is femiented by Leuconostoc dextranicum or L . citrouorum to give diacetyl (2,3-butanedione) as well as acetylmethylcarbinol and 2,3-butanediol. Some acetic and propionic acids are also produced. The important contribution of diacetyl to flavor is now well recognized and detailed information on the handling of such lactic cultures in relation to their aroma-producing capacities may be found in various textbooks on dairy microbiology (Foster et al., 1957; Hammer and Babel, 1957). The biogenesis of diacetyl has been amply discussed by Lindsay (1966).

The formation of diacetyl, however, is not limited to the leuco- nostocs only. It has been shown that certain streptococci ferment citrate, producing considerable quantities of diacetyl (Swartling, 1951). A serious drawback in the propagation of these aroma bacteria would be the formation of great amounts of acetaldehyde, which impairs flavor and contributes to the so-called “green flavor” of dairy products (Keenan et al., 1966). In pure culture experiments it was


found that the greatest amount of acetaldehyde was produced by Streptococcus diacetilactis, followed by S . cremoris and S . Zactis. The advantage of S . diacetilactis in producing diacetyl in the absence of acid-producing organisms (Pack et al., 1968) is thus overshadowed by its abundant acetaldehyde production, which has been shown to attain 6 to 10 times the level of diacetyl (Keenan et al., 1966). Accord- ing to Lindsay et al. (1965) the flavor balance of butter cultures was found to be closely related to the diacetyl: acetaldehyde ratio. Desir- able full flavored cultures exhibited ratios from 3:l to about 5:l. Streptococcus thermophilus is known to produce diacetyl in yogurt, while Lactobacillus bulgaricus has been shown to form acetaldehyde in amounts adequate for the typical yogurt flavor when cultivated in the presence of S . thermophilus (Schulz and Hingst, 1954). Acido- philus milk which is widely advocated for the therapeutic use in certain gastric disorders differs from yogurt in its lesser “buttery” and more astringent flavor. These differences are probably due to the different metabolism of L. acidophilus although the nature of the flavoring components found in acidophilus milk has not yet been worked out (Davies, 1963).

The contribution of lactic organisms and other microflora to the flavor of dairy products that are regularly aged is much more difficult to assess. Various kinds of cheeses have been analyzed for their microflora and flavoring components but it is extremely difficult to distinguish between the role of the living microbial population, the oxidative processes affecting various milk ingredients or their enzymic transformations, and the contribution to the development of cheese flavor during maturation by autolytic enzymes originating in the microflora of the product. However, it is clear that microorganisms decide the course of evolution of cheese flavor. This could be shown by the technique of Mabbitt and co-workers (1955) who obtained cheddar cheese by an aseptic procedure using 8-gluconic acid lactone as acidulant. Such cheese was devoid of any cheese flavor (Reiter et al., 1966). Although these experiments repudiate the early doubts by Mabbitt (1961) as to the contribution of microorganisms to cheddar flavor, it cannot be stated with certainty what starter cultures are most suitable. Various streptococci, such .as S . cremoris and S. lactis, are generally used in starter cultures, But lactobacilli as well as micrococci and other bacteria increase in number during maturation. Using the aseptic technique, the Reading group has demonstrated that cheeses made with starter organisms alone, had a distinct flavor although they did not develop the fullness of flavor obtained with

66 P. MAHGALITI-1 AND Y. SCHWAHTZ

ordinary cheddars. Hence, starter organisms should be considered responsible only in part for the development of cheese flavor (Reiter et al., 1967).

Undoubtedly, proteolysis has its share in the formation of cheese taste. Amino acids and their degradation products such as tyramine, cadaverine, putrescine, histamine, tryptamine, ammonia, and hydrogen sulfide, have been claimed to contribute to the flavor of cheese. Today, these substances are more considered to be of secondary importance, providing the background for the typical flavor (Reiter et al., 1967). Although rennet is the main source of proteolytic enzymes, nonstarter cheeses failed to develop any flavor resembling that of normal cheese. Hence, such proteolysis cannot be responsible for cheese flavor (Mabbitt et al., 1955). That proteolysis may play an important part in the off-flavor of milk products has been indicated by various workers. Bitter flavor has been shown to be due to “bitter” strains of S. cremoris which seem to form a peptide that causes the bitter sensation. Although this peptide is probably derived from milk casein, its mode of formation is yet unclear (Emmons et al., 1960; Czulak and Shimmin, 1961).

It is generally considered that the free fatty acids (FFA) are associated with flavor intensity of cheeses (Patton, 1963). Acetic, propionic, butyric, and higher fatty acids have been identified. Microorganisms may produce lower fatty acids from carbohydrates, but it seems more likely that the greater part of the FFA are derived from milk fat through lipolysis (Reiter et al., 1966). Although lipolytic organisms such as micrococci and gram-negative bacteria may contribute considerably to the FFA components of flavor, the weak lipolytic activities of the abundant starter organisms should not be overlooked. Since cheese flavor was shown to develop in starter cheeses even under aseptic conditions, after all the starter organisms had perished, it is evident that autolysis during ripening plays an important role in the buildup of flavor in aged dairy products. Also here the contribution of milk lipases must be tiiken into consideration (Reiter et al., 1967).

Off-flavors frequently encountered in the dairy industry, have drawn the attention of many investigators with technological and analytical interest. Gas-liquid chromatography of volatiles produced by organisms that were previously found to be related to “maltiness” has shown that malty strains of Streptococcus lactis var. maltigenes produce a variety of compounds, including 2-methylpropanol and 3-methylbutanol, but that the typical malty aroma was due principally to aldehydes, mainly 3-methylbutanal (Morgan et al., 1966). Veda-


muthu and co-workers (1966a,b), in a series of papers dealing with the problem of the so called “fruity” off-flavor of Cheddar cheese found no difference in the sugar-protein degradation potential of lactic cultures leading to normal and fruity products, respectively, but observed a considerable difference in the carbonyl production of starters. High values were found with cultures consisting of S. lactis and S. diucetiluctis that led to fruitiness, while cultures of S. cremoris and Leuconostoc sp., that produced less than 20 p.p.m. carbonyl compounds, gave normal flavor cheeses with good closed texture. Acetaldehyde, pyruvic acid, and diacetyl were the chief carbonyl compounds investigated. Thus, ripened cheeses may suffer from off- flavors similar to those of nonaged dairy products (see p. 64). A new type of off-flavor in cheese that was due to the undesirable activities of bacteria, has been recently described in Dutch Gouda cheeses (Badings et al., 1968). It was found that a microflora closely related to the group of Lactobacillus plantarum and L. casei was involved in the production of a certain phenolic off-flavor, the responsible component of which was identified as p-cresol. It was found that these undesirable organisms were derived from a contamination of inade- quately filtered rennet.

The chemistry of flavor in mold-ripened cheeses is even more complex. In addition to the processes leading to the formation of normally ripened cheeses, the metabolism of molds developing on, or within such cheeses contributes to the characteristic flavor. Roquefort, Gorgonzola, and other blue-veined cheeses are ripened with the aid of Penicillium roqueforti, that in addition to the softening of the curd and the formation of various protein breakdown products, shows a very high lipase activity that leads to the rapid production and accumulation of a number of fatty acids (mainly caprylic acid, but also capric and caproic acids). These fatty acids seem to be involved in the formation of methyl ketones, which probably are one of the most important components of cheese flavor. There have been indications with regard to the formation of such methyl ketones by the microflora involved in the ripening process of Cheddar cheese (Walker and Harvey, 1959). However, this has been later shown not to be the case, since the precursors of methyl ketones were found to be in the milk and an overall similarity could be demonstrated between the yields of methyl ketones by distillation of fat from milk and that obtained from Cheddar cheese (Lawrence, 1963; Hawke, 1966). In the case of mold-ripened cheeses, the fungal production of methyl ketones seems to be of primary importance from the organoleptic standpoint. Patton


( 1950, 1951) has isolated 2-pentanone, 2-heptanone, and 2-nonanone from mold-ripened cheese, emphasizing the importance of 2-heptanone to the characteristic cheese flavor. A comparison of the methyl ketone-producing capacity of a number of molds has been published recently by Kubeczka (1968). The formation of these ketones is supposed to be the result of the following enzymic reactions: (1) The liberation of fatty acids from milk triglycerides; (2) oxidation of the free fatty acids to a-keto acids; and (3) decarboxylation of a-keto acids to methyl ketones (Hawke, 1966). Gehring and Knight (1963) maintain that the formation of methyl ketones from fatty acids is confined to the fungal spores and is not a function of mycelial cells.

The ripening of Camembert cheese involves apparently not only the surface growth of P . camemberti, but also that of various film yeasts and Geotrichum spp., that reduce the acidity of the cheese surface before the main mold becomes established. Information on the chemical compounds involved in Camembert maturation and the specific contribution of each group of organisms to the development of the characteristic Camembert flavor is still very meager. The same may be said about microorganisms involved in the specific flavor of slimesurface cheeses, such as Brick, Limburger, and others.

D. PICKLES

The action of lactic acid bacteria on vegetable material is utilized for the purpose of preservation of nutritionally important ingredients as well as for the production of different types of foodstuffs with a special appeal to the consumer. From the microbiological point of view, this type of fermentation differs from that of the dairy industry not only in the nature of the raw materials, but also in the fact that pickling involves an environment characteristic in its high saline content.

Since the early days of food microbiology and up to recent times, shidies on the nature of cucumber pickling, sauerkraut and olive fermentation, etc., have concentrated on the microorganisms involved in the process, and the transformation of fermentable sugar into lactic acid. Good and bad fermentations have been evaluated by the appearance of the final product and its taste characteristics. In considering the place of the microorganism in the formation of the characteristic flavor of fermented plant material, we are confronted with a number of difficulties due to the complex nature of these processes. That lactic acid is not the only factor in the formation of pickle-flavor can be easily assessed by the fact that nonfermented pickles prepared with


similar amounts of such acid and salinity can be easily distinguished from fermented ones. However, during fermentation it is rather difficult to distinguish between the interaction of the microbial systems involved in the process and that of the enzymes that are present in the plant material. This difficulty is further aggravated by the fact that all vegetables employed for the lactic fermentation carry a very heavy load of microorganisms which do give way to a succession of lactic organisms, but may also affect the final organoleptic qualities of the product. In order to establish the part played by the different types of organisms, two pieces of information would be invaluable: (1) the approximate chemical composition of the flavor components of each type of fermented products, and (2) the share of each of the microorganisms involved in the lactic fermentation, when these are carried out under pure culture conditions. Although these prerequisites seem to be almost unattainable, recent studies have at least begun to con- bate upon these approaches.

An attempt to correlate the nonlactic acid components of pickles with flavor quality of the fermented product, has been made by Christensen and co-workers (1958), who analyzed the acetic acid composition of fermented vegetables. In using a setup which per- mitted pure culture fermentations, it was found that homofermentative lactics produced only minor amounts of acetic acid, independently of the sugar concentration available; while with heterofermentatives the amount of acetic acid formed during the fermentation increased with the available sugar concentration. Considerable variations have been encountered with the different strains employed. In a later paper, Pederson et al. (1962) emphasized the importance of the velocity of the fermentation process to the organoleptic qualities of the product. In a comparison of the fermentation of Yugoslavian cabbage, it was found that whole cabbage kraut was superior to that of shredded kraut, the former developing a mellow flavor, while the latter showed a more pungent, acid taste. This was probably due to the rapid fermentation which took place in the shredded cabbage, leading to higher volatile acidity. Recently Hardlicka et al. (1967) studied the formation of carbonyl compounds in kraut. Acetaldehyde and diacetyl seem to be formed during the first days of fermentation, declining thereafter.

A new approach to the study of the changes that take place during vegetable fermentations was that of Vorbeck et al. (1963) and Pederson and co-workers (1964). These workers concentrated on the behavior of the lipid fraction during fermentation. It was found that in kraut the amount of free fatty acids (FFA) increased considerably during

70 P. MARGALITH AND Y. SCtIWAHTZ

the lactic fermentation. On the other hand the amount of unsaponifi- able matter and fatty acids of both acetone soluble and insoluble lipid fractions decreased during the fermentation. A general increase in the shorter fatty acids was observed. The presence of longer-chain fatty acids in the nonesterified fatty acid fraction of the fermented material that were absent in the raw material, hus been attributed to the un- saponifiable fraction, and not only to the hydrolysis of lipids during fermentation. Although lactics are not considered lipolytic bacteria, the activity of lipolytic enzymes during autolysis may be involved. In- deed, recently Oterholm et al. (1968) have demonstrated the occur- ence of a number of lipolytic endoenzymes in many lactic acid bacteria. Also, it is not clear to what extent plant lipolytic enzymes may take part in this transformation. The chemical changes taking place during kraut fernientation have, however, not been discussed from the point of view of flavor formation or organoleptic evaluation. More pertinent data were obtained by the study of the cucumber ferinenta- tion (Pederson et al., 1964). There was a general increase in the concentration of free fatty acids, neutral fats, and unsaponifiables. Gas-liquid chromatography of the methyl esters showed that a sharp increase in linolenic acid took place in the normal product, while in that of “bloaters” an increase in oleic acid was evident. Further- more, the disappearance of tridecenoic acid from cucumbers and appearance of caproic, caprylic, and capric acids were noted. A striking characteristic of the cucumber fermentation was the decrease in the phospholipid fraction, down to 10% of the original. This, however, was not commensurate with the increase in FFA, neutral fat, and unsaponifiables. An active synthesis is, therefore, postulated to take place during the lactic fermentation. Incidentally, the breakdown of phospholipids during cucumber pickling has been noted earlier by Keil and Weyrauch (1937), who noted the accumulation of acetylcholine and lactylcholine in fermented foods and ascribed it to the activity of Bacterium acetylcholini which was later identified as a strain of Lactobacillus plantarum (Rowatt, 1948).

Although details on the chemical transformations that take place during vegetable fermentations are accumulating rapidly, information on the chemical nature of the flavor of fermented olives is still very meager. Only the off-flavor leading to the malodorous “zapatera” fermentation has received considerable attention. The first off-odor to appear in zapatera has been described as “cheesy,” developing later into a foul, fecal stench. This is accompanied by a continuous loss of acidity. Delmouzos and co-workers (1953) have shown that at least


part of the odor results from the volatile acids that develop in the brine; these include formic, propionic, butyric, valeric, caproic, and caprylic acids. In contrast, normal brines contain acetic, lactic, and sometimes succinic acids. Plastourgos and Vaughn (1957) believe that the propionic acid is produced by Propionibacterium pentosaceum and P . zeae. These organisms were shown to develop similar cheesi- ness in uitro. However, these authors believe that the propionic acid fermentation can be considered only the first stage in the malodorous fermentation, to be followed by other anaerobic organisms.

The preservation of meat products, like sausages, ham, and bacon, through the action of salts (NaCl, NaN03, and NaN02), and their subsequent maturation are generally referred to as meat curing. The carbohydrate-rich substrate of vegetable pickling is thus replaced by the proteinaceous material subjected to curing. This, however, leads to a completely different microflora, comprising a large number of genera and species (Jensen, 1954). While one of the major purposes of meat curing is the stabilization of color through the action of nitrite and the formation of nitric oxide hemoglobin, the contact of meat with such a diversified microflora, both under conditions of brine and dry curing, must affect the organoleptic qualities of the meats. Although several attempts have been made to study the effect of a number of parameters on the selective enrichment of microorganisms in brines (Deibel et al., 1961; Shank et al., 1962) comparatively little is known on the specific effect of the microorganisms that may be involved in the formation of the characteristic flavor. Only limited information on the incorporation of specific “aroma cultures” during the curing processes is available from literature. Thus, Niniivaara (1955) attempted the introduction of a Micrococcus into raw sausage. McLean and Sulzbacher (1959) reported the use of a Pseudomonas in a meat curing brine. Deibel, Wilson, and Niven (1961) used a Pediococcus as a starter in the preparation of summer sausages. A commercial preparation under the name of “Equinibe” has been marketed in France, probably a culture of Vibrio costicolus (Ribeiro, 1964) in order to improve color and flavor of ham. As Ingram (1966) pointed out, the main result of such practices was to minimize the chances of spoilage by alien species, although no high grade of flavor quality could be achieved. Knowledge on the contribution of microorganisms to the curing processes of meat is, thus, again fragmentary owing to the fact that so very little is known of the chemical entities involved in meat flavor. More recent work seems to provide new information on these aspects (Tarr, 1966).


E. ORIENTAL FOODS The current international effort to provide new foods to needy popu-

lations has drawn the attention of many workers to the vast array of fermented foods which for centuries have served many nations in the Far East. Various plant and animal materials are being used for ;he fermentative processes employed in the preparation of these foods. For an exhaustive review on the nature and mode of preparation of such foods, the reader is referred to the excellent surveys by Hessel- tine (1965), Hesseltine and Wang (1967), Amano (1961), and Saisithi et aZ. (1966). The main purpose in the fermentative production of these foods is to provide and enhance their organoleptic qualities. Fermentation adds to the flavor of these foods. Sometimes these flavors become so strong as to provide the raw material for the blending of other foodstuffs. Another advantage of the microbial process is to mask or even to destroy a number of repulsive components occurring in the raw material.

Changes in the flavor qualities due to microbial activity in a number of well known oriental foods will be briefly discussed. Fish fermentation is one of the oldest processes which varies greatly according to the type of fish employed and local practice. For example, in Thailand Nam-pla is prepared from various small fish which are mixed with salt (25%) and fermented at about 40°C in sealed tanks for about 6 months until most of the material is liquefied. The filtrate is then ripened for a number of months under the sun. The finished product is a dark brown liquid with a distinct sharp flavor (Visco and Fratoni, 1963; Saisithi et al., 1966). Clearly, the flavor of this fish sauce is the result of the microbial activity of a number of halophilic microorganisms. Among these coryneform bacteria, streptococci, micrococci, and staphylococci as well as a Bacillus sp. were isolated. Quantitative information on the occurrence of these organisms and their respective proteolytic action is still unavailable. Most of the organisms were shown to produce volatile acids. Jones (1961) claimed that the flavor of the fish sauce arises in part from glutamic and other amino acids like proline and histidine, liberated during proteolysis. Saisithi and co-workers (1966), however, believe that the distinct flavor of the fish product is a blend of organic acids and a number of amines like gluta- mine, histamine, glucosamine, and trimethylamine. Low molecular weight volatiles, such as methyl ketones and other carbonyl compounds have been suggested by others (Yanagihara et al., 1963; Yurkowski, 1965). Clearly, the relationship between the flavor of the fermented fish product, microbial activity, and the autolytic action of the fish tissues warrants more extensive investigations.


The production of soya sauce is usually preceded by the preparation of koji which serves as a starter for the main fermentation. Koji is prepared by the fermentation of wheat bran and soybean flour cultured with the mold Aspergillus oryzae or A. soya in trays or small containers. Proteolysis is promoted by the incubation at suitable temperatures. It is not clear to what extent the degree of proteolysis affects the organoleptic qualities of the final product (Sugita, 1956). Asao and Yokotsuka (1957) emphasize the importance of phenolic compounds during the koji fermentation. It was found that most of the phenolic compounds such as 4-ethylguaiacol and vanillic acid increased gradually before spore formation of the mold, while that of ferulic acid diminished and vanillin disappeared (see also phenolic compounds in Sake, p. 63).

The koji is then mixed with equal amounts of salt brine to form the mash (moromi). During a prolonged incubation period a lactic fermentation sets in, to be followed by a yeast alcoholic fermentation. A remarkable decrease in the concentration of malic and citric acids, probably due to microbial activity, has been observed during the initial stages of the mash fermentation (Ueda et al., 1958). After some aging, the liquor is separated by pressing, and constitutes the shoyu. The dark brown liquid is very salty and has a distinct sharp flavor. According to Yokotsuka (1960), the flavor of shoyu is very complex. In addition to salt, a comparatively high concentration of protein breakdown products such as peptones, peptides, and free amino acids, especially glutamic acid, may be considered as principal flavoring materials. In addition, organic bases, derived from the breakdown of nucleic acids, organic acids, such as acetic, lactic, succinic, and pyroglutamic acids, as well as a number of alcohols and esters have been mentioned to contribute to the flavor of shoyu (Shigemi and Michiyo, 1966). It is evident that most of the characterization and analysis of the distinct shoyu flavor have dealt little with the source of each flavoring component or their respective biochemical pathways. Obviously, a considerable number of these compounds are derived from the different fermentation processes in the koji or later during the fermentation of the mash. However, information on the specific action of microorganisms in this complex process is still very meager.

Recent studies on the nature of shoyu flavor have shown that a number of phenolic compounds may be considered to take part in the formation of such flavor (Yokotsuka et al., 1967a,b; Asao et al., 1967); 4-ethylguaiacol, 4-ethyl phenol, and 2-phenylethanol were shown to be involved. Torulopsis spp. (especially T. uersatilis) were


found to be responsible for the transformation of ferulic acid into alkyl phenols, while most of the yeasts isolated from the shoyu process (such as Saccharomyces rouxii) produced 2-phenylethanol (P-phenylethyl alcohol). It was found that high quality shoyu contained 0.5 to 2.0 p.p.m. 4-ethylguaiacol.

Ill. Concluding Remarks

Although this review does not pretend to be an exhaustive survey of flavor in fermented products, an attempt has been made to expose the industrial microbiologist who usually is not concerned with food production, to various aspects of taste and aroma in a number of processes that are controlled by the activities of microorganisms. Since the early days of applied microbiology until today, when practically all problems of flavor in foodstuffs are studied by means of gas-liquid chromatography techniques, a wealth of analytical data has accumulated which permit a deeper insight into the contribution of the different microbial populations to the buildup of flavor in fermented food products. In the present chapter reference has been made only to processes where microbial activity is taking place in situ during the manufacturing. In a subsequent chapter an attempt will be made to review recent advances in the fermentation of various chemical compounds which are used in concentrates or even pure form in order to enhance the flavor qualities of foodstuffs which normally are not subjected to a fermentation process during their manufacturing.


TABLE I CHEMICAL COMPOUNDS INVOLVED IN THE PRODUCTION OF

FLAVOR BY MICROORGANISM

Compound Formula

Alcohols

Ethanol

Glycerol

CH3CH20H

CH,OH I CH. OH I CH20H

n-Propanol

Isopropanol

n -Bu tanol

Isobu tanol (2-methyl- 1 -propanol)

2,3-Butanediol(2,3-butylene glycol) (dimethylene glycol)

Acetoin (Acetylmethylcarbinol)

Active amyl alcohol (2-methyl-I-butanol)

Isoamyl alcohol (3-methyl-1-butanol)

n-Hexanol

Furfuryl alcohol

Benzyl alcohol

2-Phenethyl alcohol (2-phenylethanol)

Tryptophan01

CH3CH2 CH2 OH

(CH,), CHOH

C H ~ C H I C H ~ C H ~ O H

(CH,)2CHCH,0H

CH3 CH2 (OH)CH2 (OHKH,

CHI CH(OH)COCH,

CH3CH2 CHCH,OH I CH3

(CH,),CHCH,CH,OH

CH3 CH2 CH2 CHI CH2 CHI OH

CH2CH20H

(continued)


TABLE I (Continued)

Compound Formula

Aldehydes, Ketones, and Acetals

Formaldehyde HCHO

Acetaldehyde CH3CH0

Propionaldehyde CHI CHzCHO

Acetone CH3COCH3

n-Butyraldehyde CH3CHzCHzCH0

Isobutyraldehyde (CH,)aCH2CH0

Diacetyl (2,3-butanedione) CH3 COCOCH

Methylethyl ketone CH~COCH~CHJ

2-Methyl- 1 -butanal CH3CH2CHCH0 1 CH3

n-Valerylaldehyde CH3CHaCHzCHzCHQ

lsovalerylaldehyde (3-methyl- 1 -butanal) (CHI )zCHCH2CH0

2-Pentanone

n-Hexaldehyde

2-Heptanone

2-Nonanone

Diethyl acetal (Acetal)

Ethylpentyl acetal

Ethyl, active amyl acetal

Ethyl, isoamyl acetal

FLAVOR AND MICROORGANISMS

TABLE I (Continued)

77

Compound Formula

,OCHi CH, CH(CH3 )Z

Isoamyl, pentyl acetal CH,CH, OCHZCH, CHZCHZCHS

Diactive, amyl acetal

Active amyl, isoamyl acetal

Active amyl, pentyl acetal

Diisoamyl acetal

Acids

Formic

Acetic

Propionic

Lactic

Pyruvic

n-Butyric

Isobutyric

Succinic

Malic

n-Valeric

lsovaleric

2-Me thylbutyric

OCH1CHCH,CH3

CH,CH( LH, OCH, CHCH, CH,

I CH,

OCH, CHCH,CH, / I

CH,CH CH3 ‘OCH~CH,CH(CH,),

OCH, CHCH, CH,

OCH2CH, CH,CH,CH,

/ I CH,CH, CH3

,OCHzCHiCH(CH,), CHJCH

\OCH~CH,CH(CH,),

HCOOH

CH,COOH

CH,CH,COOH

CH,CH(OH)COOH

CH,CG€OOH

CH3CH2CH,COOH

(CH, ),CHCOOH

HOOCCH,CH,COOH

HOOCCH, CH(0H)COOH

CH,CH,CH, cn, coon

(CH, IICHCH ,COOH

CH3CHaCH(CH3 )COOH

(continued)


TABLE I (Continued)

Compound Formula

Acetolactic

(I- Ketoglutaric

nCaproic

CHI COCH, CH(0H)COOH

HOOCCH, CH, COCOOH

CH, (CH,)4COOH

Citric

Isocaproic

2-H ydroxyisocaproic

n-Heptylic

nCaprylic

Pelargonic

nCapric

9-Decenoic

Lauric

Tridecanoic

Myristic

Palmitic

Oleic

Linoleic

Linolenic

Amino acids and amines

Trimethylamine

Putrescine

Cadaverine

Clutarnic acid

Glutainine

HOOCCH, C(OH)CH1 COOH I COOH

(CH3 ), CHCH~CH, COOH

(CH, ), CHCH,CH(OH)COOH

CH3(CHl)s COOH

CH3(CH,)6COOH

CH,(CH,),COOH

CH,(CHl ),COOH

CH,=CH(CH, ),COOH

CH, (CH,)ioCOOH

CH,(Cli,), I COOIi

CH,(CHl),2COOH

CH,(CH2 ),,COOH

CH~(CH~)~CH=CH(CHZ ),COOH

CH,(CH, )a (CH,CH=CH), (CH, ),COOH

CH3 (CH2CH=CH)~(CH1 ),COOH

NH2CH,(CHz)iCHzNHa

NH,CH,(CH1)3CH1NH2

HOOC(CHz)2CH(NH, )COOH

HOOC(CH, ),CH(NH, )CONHI

FLAVOR AND MICROORGANISMS

TABLE I (Continued)

Formula

Pyroglutamic acid

Proline

Histidine

Histamine

Glucosamine

Tryptamine

Eaters

Ethyl formate

Ethyl acetate

2-Phenylethyl acetate

1,3-Propanediol monoacetate

lsoamyl acetate

H omcooH H nco0,,

HO Ho OH OH

H NH,

a CH,CH,NH,

HCOOCH2CH3

CH3COOCH2CH3

CH~COOCHZCH,C,HS

CH,COOCH,CH,CHzOH

CH3COOCH2 CHzCH(CH3 )i

(continued)

P. MARGALITH AND Y. SCIiWARTZ

TABLE I (Continued)

Compound Formula

Ethyl lactate

Ethyl Chydroxybutyrate

Diethyl succinate

Diethyl malate

Ethyl pyrodutamate

2-Phenylcthyl caproate

Acetylcholine

Lactylcholine

Phenol compounds

p-Cresol

4-Ethylphenol

CHI CH(OH)COOCH, CH,

(HO)CHtCH, CHI COOCH, CHS

CH2COOCHzCH3 I CH2COOCHzCH3

(OH)CHCOOCH2 CH3 I Cl12C00Cti,CH3

H

O r a c H 3

CHJ(CH2),COOCHlCH,C~H~

CH3COOCHzCHz N(CH3 +

+ CH3 CH(0H)COOCHZ CHI N(CH3 )3


TABLE I (Continued)

Compound Formula

p-Hydroxybenzaldehyde

p-Hydroxybenzoic acid

Tyrosol

4Methylguaiacol

4-Ethylguaiacol

on I

Q

on I

Q COOH

OH I

OH

Qocn3

OH

3”:”. (continued)

P. MARGALITH AND Y. SCHWARTZ

TABLE I (Continued)

Compound Formula

4-Vinylguaiacol

Vanillin

Vanillic acid

Ferulic acid

OH I

CH=CIla

OH I

CHO

COOH

OH I

qoCH3 CH=CHCOOH

Sulfur compounds

Hydrogen sulfide

Thioformaldehyde

Dithiofonnaldehyde

Thioacetone

Ethanethiol

Has

HCHS

(HCHSh

CH3CSCH3

CH3CH2 SH


REFERENCES

Amano, K. (1961). Fish Nutr. Fishing News (London), 180-200. Amerine, M. A. (1954). Aduan. Food Res. 5,353-510. Amerine, M. A. (1958). A p p l . Microbiol. 6, 160-168. Amerine, M. A., and Cmess, W. V. (1960). “The Technology of Wine Making.” Avi

Amerine, M. A., Pangborn, R. M., and Roessler, E. B. (1965). “Principles of Sensory

Amoore, J. E. (1952). Perfumery Essent. Oil Record 43,321-323. Amoore, J. E., and Venstrom, D. (1966).]. Food Sci. 31,118-129. Amoore, J. E., Johnston, J. W., Jr., and Rubin, M. (1964). Sci. Am. 210,42-49. Antoniani, C. E. (1961). Proc. European Brewery Conu. Congr. Vienna, pp. 216-223. Arkima, V., and Shito, E. (1963). Proc. European Brewery Conu. Congr. Brussesls,

Arthur, D., Little, Inc. (1958). “Flavor Research and Food Acceptance.” Reinhold, New

Asao, Y., and Yokotsuka, T. (1957). Bull. Agr. Chem. Soc.]apan. 32,622-628. Asao, Y., Sakasai, T., and Yokotsuka, T. (1967). Nippon Nogeikagaku Kaishi 41,428-433. Ayrapaa, T. (1963). Proc. European Brewery Conu. Congr. Brussels, pp. 276-287. Badings, H. T., Stadhouders, J., and Van Dum, H. (1968). 1. Dairy Sci. 51,31-35. Bavisotto, V. S., and Roch, L. A. (1959). Proc. Am. S O C . Brewing Chemists, pp. 63-75. Bavisotto, V. S., and Roch, L. A. (1960). Proc. Am. S O C . Brewing Chemists,pp. 101-112. Beidler, L. M . (1966). In “Flavor Chemistry” (R. F. Could, ed.), p. 4. Am. Chem. Soc.,

Benvegnin, L., Capt, E., and Piquet, G . (1951). “Traite de Vinification,” 2nd ed., pp.

Berg, C. (1953). Physiol. Reu. 33, 145-189. Bobadilla, G . F. (1943). Agricultura (Louuain) 12, 203-207. Boyd, W. C., and Matsubara, S. (1962). Science, 137, 669. Brenner, M. W., Blick, S. R., Frenkel, G., and Siebenberg, J. (1963). Proc. European

Brumsted, D. D., Lautenbach, A. F., and West, D. B. (1965). Proc. Am. S O C . Brewing

Bujas, Z. (1934). Compt. Rend. SOC. Biol. 116, 1304-1309. Cappucci, C. (1948). Riu. Viticolt. Enol. 1, 386-388. Carlin, G . T. (1958). Proc. Am. SOC. Bakery Engr. pp. 56-63. Carlin, G . T. (1959). Proc. Am. S O C . Bakery Engr. pp. 136-147. Castor, J. G. 3., and Archer, T. E. (1957). Appl. Microbiol. 5, 56-60. Caul, J . F., and Raymond, S. A. (1964). Food Technol. 18,353-357. Champagnat, A., Vernet, C., Lain&, B., and Filosa, J. (1963). Nature 197, 13-14. Christensen, M. D., Albury, M. N., and Pederson, C. S. (1958). A p p l . Microbiol. 6,

Cole, E. W., Hale, W. S., and Pence, J. W. (1962). Cereal Chem. 39, 114-122. Cole, E. W., Hale, W. S., and Pence, J. W. (1963). Cereal Chem. 40, 260-265. Crowell, E. A., and Guymon, J. F. (1963). Am. 1. End. Viticult. 14, 214-222. Czarnecki, H. T., and Van Engel, E. L. (1959). Brewers Dig. 34, 52. Czulak, J., and Shimmin, P. D. (1961). Australian J . Dairy Technol. 16, 96-98. Davies, J. G. (1963). Prog. Ind. Microbiol. 4, 95-136. Davies, J. T., and Taylor, F. H. (1954). Nature 174,693-694.

Publ., Westport, Connecticut.

Evaluation of Food.” Academic Press, New York.

pp. 268-275.

York.

Washington, D.C.

324-328. Librairie Payot, Lausanne.

Brewery Conu. Congr. Brussels, pp. 233-246,

Chemists, pp. 142-145.

316-318.

84 P. MAHGALITII AND Y. SCHWARTZ

DeClerk, J. (1958). “A Textbook of Brewing.” Chapxiian and Hall, London. Deibel, R. H., Wilson, G. D., and Niven, C. F. (1961). Appl. Microbiol. 9, 239-243. Delmouzos, J. G., Stadtinan, F. H., and Vaughn, R. H. (1953). J . Agr. Food Chem. 1,

Drews, B., Specht, H., and Barwald, G . (1964). Monatsschr. Brauerei 17, 101-116. Drews, B., Specht, H., and Schwartz, E. (1965). Monatsschr. Bruuerei 18,240-242. Duncan, C . J. (1963). J . Theoret. Biol. 5 , 114-126. Emmons, D. B., McCuyan, W. A., and Elliott, J. A. (1960).J. Dairy Sci. 43,862. Enebo, L. (1957). Proc. Europeun Brewery Conu. Congr., Copenhagen, pp. 370-376. Farice, W. J. (1960). “The Students Technology of Breadmaking and Flour Confec-

Fiechter, A. (1967). Chimia 21, 501-508. Fornachon, J. C. M. (1957). Austrulian]. Appl. Sci. 8, 120-129. Fornachon, J. C. M., and Lloyd, B. (1965). J . Sci. Food Agr. 16, 710-716. Foster, E. M., Nelson, F. E., Speck, M. L., Doetch, R. N., and Olson, J. C. (1957).

“Dairy Microbiology.” Prentice Hall, Englewood Cliffs, New Jersey. Galetto, W. G., Webb, A. D., and Kepner, R. E. (1966). A m . ] . Enol. Viticult. 17, 11-19. Gehring, R. F., arid Knight, S. G . (1963).Appl. Microbiol. 11,166-170. Gentilini, L., and Cappelleri, G. (1959). Ann. Sper. Agrar. (Rome) 13,289-306. Gilliland H. B. (1961). J . Inst. Brewing 67, 257-261. Gilliland, R. B., and Harrison, C . A. F. (1966). /. Appl. Bacteriol. 29,244-252. Gjertsen, P., Undstrup, S., and Trolle, B. (1964). Monutsschr. Brauerei 17, 232-234. Guymon, J. F. (1966). Develop. Ind. Microbiol. 7, 88-96. Guymon, J. F., and Crowell, E. A. (1965). Am. J . End. Viticult. 16, 85-91. Guymon, J. F., and Heitz, J. E. (1952). Food Technol. 6, 359-362. Haimrier, B. W., and Babel, F. J . (1957). “Dairy Bacteriology,” 4th ed. Wiley, New York. Hardlicka, H., Curda, D., and Pavelka, J. (1967). Sh. Vysoke Skoly Chem. Technol.

Harold, F. V., Hildebrand, R. P., Morieson, A. S., and Murray, P.], (1961).J. Inst. Rrew-

Hartl, A. (1964). Brauwelt 104, 1117-1127. Ilartong, B. D. (1963). Proc. European Brewery Conu. Congr. Brussels, pp. 528-529. Hashimoto, N., and Kuroiwa, Y . (1966). Proc. Am. SOC. Brewing Chemists, pp. 121-130. Hawke, 3. C. (1966). J . Dairy Res. 33, 225-243. Hesseltine, C. W. (1965). Mycologia 57, 149-197. Hesseltine, C. W., and Wang, H. L. (1967). Biotechnol. Bioeng. 9,275-288. Hickinbotham, A. R., and Ryan, V. J. (1948). Australian Chem. Inst. J . G Proc. 15,

Hind, H. L. (1950). “Brewing, Science and Practice.” Chapman and Hall, London. Hinreiner, E., Filipello, F., Berg, H. W., and Webb, A. D. (1955). Food Technol. 9,

Hornstein, I., and Teranishi, R. (1967). Chem. Eng. News 45, 92-108. I-Iough, J. S., and Stevens, R. (1961). J . Inst. Brewing 67, 489. IIowell, W. H. (1922). “Tcxtbook of Physiology,” 8thed., p. 295. Saunders, Philadelphia,

Hudson, J. R., and Stevens, R. (1960). J . Inst. Brewing 66, 471-474. Hunter, I. G., NG, H., and Pence, J. W. (1961).]. Food Sci. 26, 578-580. Ingram, M. (1966). J . Appl. Bacteriol. 29, 217-221.

333-334.

tionery.” Routledge and Kegan Paul, London.

Praze Potravinarloke Technol. 15,51-54.

ing 67, 161-172.

89-100.

489-490.

Pennsylvania.


Ingraham, J. L., Vaughn, R. H., and Cooke, G. M. (1960). Am. J . E n d . Viticult. 11,l-4. Ingraham, J. L., and Guymon, J. F., and Crowell, E. A. (1961). Arch. Biochem. Biophys.

Jenard, H., and Devreux, A. (1964). Echo Brass. 20, 1119. Jensen, L. B. (19.54). “Microhiology of Meats,” 3rd ed. Garrard Press, Champaign,

Jerchel, D., Flesch, P., and Bauer, E. (1956). Ann. Chem. 601, 40-60. Johnson, J. A., Miller, B. S., and Curnutte, B. (1958). J . Agr. Food Chem. 6 , 384-387. Jones, N. R. (1961). Proc. Flavor Chem. Symp. pp. 61-81. Campbell Soup Co. Camden.

Joslyn, M. A., and Dunn, R. (1941).J. Am. Chem. SOC. 60, 1137-1141. Keenan, T. W., Lindsay, R. C., Morgan, M. E., and Day, E. A. (1966). J . Dairy Sci. 49,

Keil, W., and Weyrauch, L. (1937). Zentr. Bakteriol. Parasitenk. Abt. II.97,90-94. Kepner, R. E., and Webb, A. D. (1956). Am.] . Enol. 7,8-18. Kepner, R. E., Strating, J., and Weurman, C. (1963). J . Znst. Brewing 60, 399-405. Kohn, F. E., Wiseblatt, L., and Fosdick, L. S. (1961). Cereal Chem. 38,165-169. Komoda, H., Mano, F., and Yamada, M. (1966). Nippon Nogeikagaku Kaishi 40, 127-

Kubeczka, K. H. (1968). Arch. Mikrobiol. 60, 139-159. Kuninaka, A. (1966). Adwan. Chem. Ser. 56,261-274. Kunkee, R. E. (1967a). Am. J . Enol. Viticult. 18, 71-77. Kunkee, R. E. (1967b). Advan. Appl. Microbiol. 9, 235-279. Kurtzman, C. H., and Sjostrom, L. B. (1964). Food Technol. 18,1467-1469. Lafon-Lafourcade, S., and Peynaud, E. (1966). Ann. Znst. Pasteur 110,776-778. Lambion, R., and Meskhi, R. (1957). Rew. Ferm. Znd. Aliment. 12, 131-144. Lawrence, R. C. (1963). J . Dairy Res. 30, 161-170. Lawrence, W. C. (1964). Wallerstein Lab. Commun. 27, 123-152. Lewin, L. M., and Smith, E. J. (1964). Nature 203, 867-868. Linko, Y. Y., Miller, B. S., and Johnson, J. A. (1962). Cereal Chem. 39,263-272. Lindsay, R. C. (1966). In “Chemistry and Physiology of Flavors” (H. W. Schultz, E. A.

Lindsay, R. C., Day, E. A., and Sandine, W. E. (1965)J. Dairy Sci. 48,863-869. Lipis, B. V. , and Mamakova, Z. A. (1963). Vinodelie i Vinogradarstvo SSSR. 23, 7-11. Lukas, C., cited by Schanderl (1959). “Mikrobiologie des Mostes und Weines,” p. 101.

Mahbitt, L. A. (1961). J. Dairy Res. 28, 303-318. Mabhitt, L. A, , Chapman, H. R., and Berridge, N. J. (1955). J. Dairy Res. 22, 365-373. McLean, R. A., and Sulzbacher, W. L. (1959). Appl. Microbiol. 7, 81-83. Marcilla, J. A,, Alas, G., and Feduchy, E. (1936). Anal. Centro. Inwest. Vinicolas 1,

Masschelein, C. A., Jenard, H., Jeunehomme-Ramos, C., and Devreux, A. (1965). Proc.

Matz, S. A. (1960). “Bakery Technology and Engineering.” Avi Publ., Westport,

Miller, B. S., and Johnson, J. A. (1958). Wallerstein Lab. Commun. 21,115-132. Miller, B. S., Johnson, J. A., and Robinson, R. J. (1961). Cereal Chem. 38, 507-515. Moncrieff, R. W. (1951). “The Chemical Senses,” 2nd ed. Leonard Hill, London.

95,169-175.

Illinois.

New Jersey.

10-14.

134.

Day, and M. Lihbey, eds.), p. 315. Avi Publ. Westport, Connecticut.

Ulmer, Stuttgart.

1-230.

European Brewery Conu. Congr. Stockholm, pp. 209-22.

Connecticut.


Moncrieff, R. W. (1966). “Odour Preferences.” Leonard Hill, London. Morgan, M. E., Lindsay, R. C., Libbey, L. M., and Pereira, R. L. (1966).]. Daily Sci.

Niniivaara, F. P. (1955). Acta Agral. Fennica 84, 1-128. Nordstroin, K. (1965). Proc. European Brewery Conu. Congr. Stockholm, pp. 195-198. Ochoa, S. (1951). Physiol. Reu. 31, 56-106. Omelianski, V. L. (1923). J . Bacteriol. 8, 393-419. Omori, T., Yamamoto, A., and Yasui, H. (1968). Agr. Biol. Chem. 32, 539-548. Oterholrn, A., Ordal, J., and Witter, L. D. (1968). Appl. Microhiol. 16,524-527. Ough, C. S., and Amerine, M. A. (1960). Food Technol. 14, 155-159. Ough, C. S., and Amerine, M. A. (1967). Am. 1. E n d . Viticult. 18, 157-164. Owades, J , L., Maresca, L., and Rubin, G. (1959). Proc. Am. Soc. Brewing Chem. pp.

Owaki, K. (1967). NipponJozo Kyokai Zasshi 62,841-853. Pack, M. Y., Vedamuthu, E. R., Sandine, W. E., Elliker, P. R., and Leesment, H. (1968).

Pangborn, R. M. (1961). J. Food Sci. 26,648-655. Patton, S . (1950).]. Dairy Sci. 33,680-684. Patton, S . (1951). Food Eng. p. 93. Pattorl, S. (1963).]. D u l y Sci. 46, 856-858. Pederson, C. S., Niketic, G., and Albury, M. N. (1962). Appl. Microbiol. 10, 86-89. Pederson, C. S., Mattick, L. R., and Butts, R. M. (1964). Appl. Microbiol. 12, 513-516. Peynaud, E., and Guimberteau, G. (1962). Ann. Technol. Agr. 11,85-105. Peynaud, E., Lafoii-Lafourcade, S., and Cuimberteau, G. (1966). Am. /. Enol. Viticult.

Peynaud, E., Lafon-Lafourcade, S., arid Guimberteau, G. (1967). Antonie Van Leeu-

P f a h a n n , C. (1964). Am. Scientist 52, 187-206. Pfenninger, H. B. (1963). Proc. European Brewey Conu. Congr. Brussels, pp. 257-267. Pilone, G. J., and Kunkee, R. E. (1965). Am.J. Enol. Viticult. 16,224-230. Pilone, G. J., Kunkee, R. E., and Webb, A. D. (1966). Appl. Microbiol. 14, 608-615. Plastourgos, S., and Vaughn, R. H. (1957). Appl. Microbiol. 5,267-271. Pollard, A., Kieser, M. E., and Beech, F. W. (1966). 1. Appl. Bacteriol. 29, 253-259. Rdler , F. (1958). Arch. Mikrobiol. 31, 224-230. Rankine, B. C. (1955). A m . ] . Enol. Viticult. 6, 11-15. Rankine, B. C. (1963). J . Sci. Food Agr. 14, 79-91. Rankine, B. C. (1967). J. Sci. Food Agr. 18, 583-589. Reiter, B., Fryer, T. F., Sharpe, M. E., and Lawrence, R. C. (1966).]. Appl. Bacteriol.

Reiter, B., Fryer, T. F., Pickering, A., Chapman, H. R., Lawrence, R. C., and Shave,

Rentschler, H. (1951). Schweiz. 2.-Obst. Weinbau 60, 197. Ribkreau-Gayon, J., and Peynaud, E. (1946). Cornpt. Rend. 222,457-458. Ribereau-Cayon, J., and Peynaud, E. (1961). “Traite d’uenologie,” Vol. 2. Librairie

Ribereau-Gayon, J., and Peynaud, E. (1964). “Traite d’oenologie,” Vol. 1. Librairie

Ribeiro, A. M. (1964). Ann. Inst. Pasteur 15, 99. Rice, A. C. (1965). A m . ] . End . Viticult. 16,62-68.

49,15-18.

22-26.

I . Dairy Sci. 51, 339-344.

17,302-307.

wenhoek J. Microbiol. Serol. 33,49-55.

29,231-243.

M. E. (1967). 1. Dairy Res. 34, 257-272.

Polytcch. CIL Bkrmger, Paris.

Polytech. Ch. Beranger, Paris.


Robinson, R. J., Lord, T. H., Johnson, J. A., and Miller, B. S. (1958). Cereal Chem. 35,

Rodopulo, A. K., Levchenko, T. N., Pisarnitskii, A. F., Bezzubov, A. A., and Martakov,

Rowatt, E. (1948). J. Gen. Microbiol. 2, 25-30. Saisithi, P., Kasemsarn, B., Liston, J., and Dollar, A. M. (1966).J. Food Sci. 31,105-110. Sandegren, E., and Enebo, L. (1961). Wallerstein Lab. Commun. 24,269-279. Schanderl, H. (1943). D m t . Weinztg. 80, 43-46. Schanderl, H. (1959). “Die Mikrobiologie des Mostes und Weines,” Vol. 2. E. Ulmer,

Schulz, M. E., and Hingst, G . (1954). Milchwissenschaft 9, 330-336. Shank, J. L., Silliker, J. H., and Harper, R. H. (1962). Appl. Microbiol. 10, 185-189. Shigemi, M., Michiyo, M. (1966). Hakko Kogaku Zasshi 44,476-482. Skovholt, 0. (1964). In “Wheat Chemistry and Technology” (I. Hlynka, ed.), pp. 573-

Smith, D. E., and Coffman, J. R. (1960). Anal. Chem. 32, 1733-1737. Solms, J. (1967). Chimia 21, 169-175. Staudenmayer, T. (1961). Z. Lebensmitt. Untersuch. Forsch. 115, 16-19. Steinhardt, R. G., Calvin, A. D., and Dodd, E. A. (1962). Science 135, 367-368. Steir, E. F., Sawyer, F. M., and Fergenson, P. E. (1967). Food Technol. 21,1627-1630. Strating, J., and Venema, A. (1961). J . Inst. Brewing 67, 525-528. Sugita, N. (1956). I . Ferment. Technol. 34, 607-612. Swartling, P. (1951). J . Dairy Res. 18, 256. Tabachnick, J., and Joslyn, M. A. (1953). J. Bacteriol. 65, 1-9. Tanner, F. W. (1944). “The Microbiology of Foods,” 2nd ed., pp. 694-695. Garrard

Tarr, H. C. A. (1966). Aduan. Chem. Ser. 56,190-202. Thoukis, G., Ueda, M., and Wright, D. (1965). Am. J. End. Viticult. 16, 1-8. Uchimoto, D., and Cruess, W. V. (1952). Food Res. 17,361-366. Ueda, R., Nagi, S., Kambara, K., and Moriguchi, S. (1958). J . Ferment. Technol. 37,

Van Wyk. C. J., Kepner, R. E., and Webb, A. D. (1967).J. Food Sci. 32,664-668. Vaughn, R. H., and Tchelistcheff, A. (1957). Am.J. E n d . 8,7479. Vedamuthu, E. R., Sandine, W. E., and Elliker, P. R. (1966a).]. Dairy Sci. 49,144-150. Vedamuthu, E. R., Sandine, W. E., and Elliker, P. R. (1966b).J. Dairy Sci. 49,151-157. Venezia, M., and Gentilini, L. (1941). Ann. Staz. Sper. Viticolt. Enol. (Conegliano)

Visco, S., and Fratoni, A. (1963). S y n i p . Siibsttrticrs Etrnngkres Aliitietits, Vol. 6 , 111).

von Skramilk, E. (1921). Z. Sinnesphysiol. 53B, 36-78. Vorbeck, M. L., Albury, M. N., Mattick, L. R., Lee, F. A., and Pederson, C. S. (1963).

Wahab, A., Witzke, W., and Cniess, W. V. (1949). Fruit Prod. J . 28, 198-200, 202-219 Wagner, J. R., Titus, D. S., and Schade, J. E. (1963). Food Technol. 17,730-735. Walker, J. R. L., and Harvey, R. J. (1959). J . Dairy Res. 26, 265-272. Webb, A. D. (1967). Biotech. Bioeng. 9, 305-319. Webb, A. D., and Ingraham, J. L. (1963). Aduan. Appl. Microbiol. 5,317-353. Webb, A. D., and Kepner, R. E. (1957). Food Res. 22,384-395. Webb, A. D., and Kepner, R. E. (1961). Am. J . Enol. Viticult. 12,5149.

295-305.

A. A. (1967). Prikl. Biokhim. Mikrobiol. 3, 614-621.

Stuttgart.

589. Am. Assoc. Cereal Chemists, St. Paul, Minnesota.

Press, Champaign, Illinois.

94-104.

10,282-309.

505-525. Ministerio de Agricultura, Madrid.

J. Food Sci. 28,495-502.

88 P. MARGALITH AND Y. SCHWAHTZ

Webb, A. D., and Kepner, R. E. (1962). Am. J. Enol. Viticult. 13, 1-14. Webb, A. U., Kepner, R. E., and Galetto, W. G . (1964). Am. J. Enol..ViticuZt. 15, 1-10. Webb, A. D., Kepner, R. E., a i d Galetto, W. G. (1966). A m . J. Enol.’Viticult. 17, 1-10. Webb, A. D., Kepner, R. E., and Maggiora, L. (1967). Am. J. Enol. Viticult. 18,190-199. West, D. R., Evans, R. F., and Becker, K. (1951). PTOC. Am. Soc. Brewing Chem. Chicago,

West, L). B., Lautenbach, A. E., arid Becker, K. (1952). Proc. Am.. Soc. Brewing Chem.

White, J. (1954). “Yeast Technology.” Wiley, New York. Whiting, G. C., and Coggins, H. A. (1960). J. Sci. Food Agr. 11, 337-344. Wiseblatt, L. (1960). Cereal Chem. 37, 734-739. Wisehlatt, L., and Kohn, F. E. (1960). Cereal Chem. 37, 55-66. Wiseblatt, L., and Zouniut, N. F. (1963). Cereal Chem. 40, 162-169. Woll, E. (1955). Wein-Wiss. Beil. Fachz. Deut. Weinhau 9,l-4. Yamamoto, A. (1961). Nippori Nogeikagaku Kaishi 35, 711-715. Yarrianioto, A,, Sasaki, K., and Saruiio, H. (1961). Nippon Nogeikagaku Kaishi 35,

Yamamato, A., Omori, H., and Yasui, H. (1966). Nippon Nogeikagaku Kaishi 40, 152-

Yariagilrara, T., Komoda, H., Yoneyania, H., and Yarnada, M. (1063). Slzokuhin Eisei-

Yokotsuka, T. (1960). Aduan. Food Hes. 10, 75-134. Yokotsuka, T., Sakasai, T., and Asao, Y. (1967a). Nippon Nogeikagaku Kaishi, 41,

Yokotsuka, T., Asao, Y., and Sakasai, T. (1967b). N i p p o n Nogeikagaku Kaishi, 41,

Yosliizawa, K. (1964). A ~ T . B i d . Chem. 28,279-285. Yurkowski, M. (1965). J. Fisheries Res. Board Can. 22,17-27.

p. 51.

T ~ r o ~ ~ t o , pp. 81-88.

7 15-7 19.

160.

gaku Zasshi 4,348-353.

428-434.

442-447.

Documents

[Advances in Applied Microbiology] Volume 12 || Flavor and Microorganisms