Editor's note:This article is part of a semi-regular series of research briefs sponsored by the Department of Viticulture and Enology at UC Davis.

This month, I will talk about a whole chapter in a Microbiology book, "Bacteria Important During Winemaking," by James Osborne and Charles Edwards, in Advances in Food and Nutrition Research. S. Taylor (Ed.) 50:139-177. Here, the authors present a very fine review of each of the main microorganisms that live in wine. As in a Gary Larson cartoon, we are offered a pretty good slice of their lives; for example, how they are dressed (Gram+ or Gram-), how tall or fat they are (filaments, bacillus, rods, spheres), their eating habits (homofermentative, heterofermentative), where they like to hang out (juice, wine) or what they like to produce (acetic acid, ethyl acetate, lactic acid, SO₂, polysaccharides). Chemical compound names aside, the authors' simple style translates into a practical and very powerful text for the winemaker.

In their introduction, the authors explain the structure of the chapter. Many bacterial species are present during fermentations, and the extent of growth of each is going to determine the type and amount of compounds that will contribute to the wine's final aromas and tastes. Because these organisms have different growth requirements and tolerances, they are going to dominate at different times during the course of the fermentation. In general, only a few species are able to grow in grape juice and wine, which are considered harsh environments (low pH, low oxygen, high alcohol, high osmotic pressure). These species involve just a handful of genera: Acetobacter, Gluconobacter, Lactobacillus, Oenococcus and Pediococcus. Some of them, like the malolactic bacteria (Oenococcus), will improve wine quality; others will tend to cause spoilage (Acetobacter, Gluconobacter, Lactobacillus, Pediococcus); and still others will tend to produce compounds that are a public health concern (some Lactobacillus, Oenococcus). Finally, these bacteria will inevitably interplay with the other actors onstage, like the wine yeast Saccharomyces, something that can be beneficial or detrimental, depending on the species involved.

Acetobacter and Gluconobacter

These are normally referred to as Acetic Acid Bacteria (AAB). They are both Gram- (coated by few layers of peptidoglycan that do not absorb the Gram dye), aerobic rods, and are the microorganisms involved in commercial vinegar production--and unfortunately in commercial wine too. Gluconobacter lives in grapes and musts, but disappears as alcoholic fermentation begins. As for Acetobacter, more alcohol-tolerant, it survives during the entire fermentation. Damaged grapes can have their sugars metabolized by yeast present on the clusters, producing alcohol, which is, in turn, oxidized by these bacteria to acetic acid. Because of their need for oxygen, the survival of these AAB in the winery may be due to exposure of wine to air during pumping and transfer operations. Besides O₂, their growth is also very influenced by pH; and populations were shown to decrease from 10⁵ cfu/ml to 10² cfu/ml if the pH were reduced from 3.7 to 3.5 (cfu/ml stands for "colony forming units per millimeter" of liquid culture, a common way of measuring microbial concentration). Another difference between Gluconobacter and Acetobacter is the pathways they use to metabolize glucose. Besides glucose, both are able to use other carbohydrates present in wine (fructose, galactose, arabinose, mannose, etc.).

Acetic acid bacteria can cause wine spoilage by producing: 1) excessive acetic acid (legal limit is 1.2-1.4 g/l, but they can easily produce 1.5-3.75 g/l); 2) ethyl-acetate (flavor threshold is only 10 mg/l, but the bacteria can produce up to 140 mg/l); 3) acetaldehyde, which tends to bind and deactivate SO₂; 4) dihydroxyacetone, due to glycerol metabolism, which can give a "sweet/esterish" note to the wine, and also binds SO₂; and finally, 5) acetoin, a product of lactate metabolism, which is a precursor of diacetyl ("butter").

Lactobacillus

These are Gram+, microaerophilic bacteria (liking slight amounts of oxygen) ranging in shape from rods to coccobacilli (short rods to elongated spheres). Lactobacillus, along with Oenococcus and Pediococcus, are referred to as Lactic Acid Bacteria (LAB). Unlike the previous group, Lactobacillus have complex nutritional requirements (besides carbohydrates, they require amino acids, nucleic acids, vitamins and fatty acids). This is called being "fastidious." Regarding their hexose metabolism, Lactobacillus are divided into: 1) homofermenters and 2) heterofermenters. Homofermentive lactobacilli convert glucose to lactic acid without the production of CO₂. Heterofermentive lactobacilli go a little further, metabolizing hexoses into a "heterogeneous" amount of compounds, like lactic acid, CO₂, ethanol and acetate. The occurrence of Lactobacillus in wine is highly pH- and ethanol-dependent. In high-pH wines (greater than 3.5), Lactobacillus will dominate (at lower pH values, Oenococcus oeni will). Ethanol tolerance varies among Lactobacillus species. For example, L. plantarum ceases growth at 5 to 6 percent ethanol, whereas L. casei and L. brevis are very alcohol-tolerant, and are the ones used to induce malolactic fermentation (MLF).

Lactobacillus can damage wines by producing: 1) excessive acetic acid (for example, L. kunkeei can produce as much as 3-5 g/l); and 2) stuck or sluggish fermentations; 3) Some winemakers have observed rapid spoilage by organisms dubbed "ferocious" lactobacillus in wines that had not been treated with SO₂ and whose initial pH was higher than 3.5; finally, 4) hetero-fermentative lactobacilli have been associated with the "mousy" defect in wines. (For more on "mousy" defects in wines, see Bibiana Guerra's article, "Recent Research: Mousy Off-Flavors" in the June issue of Wine Business Monthly.)

Oenococcus

The main agent of MLF used to be classified as Leuconostoc oenos (1967), but after observing that it possessed characteristics distinct from other Leuconostoc species, it was reassigned to a new genus, Oenococcus (1995). Strains of Oenococcus oeni are Gram+, facultative anaerobes (normally use oxygen but can also live without it), ellipsoidal to spherical cells that appear in pairs or chains. Similar to Lactobacillus, they have rich, medium requirements. But unlike Lactobacillus, they are all heterofermentative (convert glucose to equal molar amounts of lactic acid, CO₂ and ethanol or acetate).

By converting malic acid (tricarboxylic or possessing three acid groups) into lactic acid (dicarboxylic), O. oeni may be desirable because it reduces a wine's acidity and contributes new flavors and aromas. Although selected Lactobacillus strains can be inoculated, O. oeni is the primary species used commercially to conduct MLF because of its acid tolerance and the flavor profile produced. Exactly when is the best time to inoculate O. oeni is still a matter of debate among researchers. The authors point out that problems that have been observed when inoculation took place, before completion of alcoholic fermentation (excessive VA, sluggish fermentations) are likely due to incompatibilities between specific strains. Since energy-rich compounds (like pyruvate) are not formed in the conversion from malic to lactic acids, what the bacteria were actually gaining out of malolactic fermentation has puzzled researchers for quite a while. That is, until 1989, when it was shown that malolactic fermentation does yield ATP (adenosine tri-phosphate), a form of metabolic energy that the bacteria can store and spend at a later time.

Besides deacidification, some other sensory impacts of MLF, and some of Oenococcus oeni, include: 1) Production of diacetyl (or 2,3-butanedione), imparting "buttery" or "nutty" aromas. This compound has sensory thresholds of 0.2 mg/L in Chardonnay, 0.9 mg/L in Pinot Noir and 2.8 mg/L in Cabernet Sauvignon. When the compound exceeds 5 mg/L, it is considered a spoilage. 2) Production of esters, many of which are responsible for a pleasant "fruity" nose. 3) Liberation of monoterpenes. These compounds are often present in grapes as non-volatile, flavorless forms, but the ß-glucosidase activity of O. oeni can free the volatile free form. 5) O. oeni can also metabolize acetaldehyde, as well as other aldehydes, to produce ethanol and acetic acid. Depending on the case and extent, this can be desirable (acetaldehyde causes an off-aroma) or undesirable (acetaldehyde also plays a role in color development). 5) Finally, MLF increases body or mouthfeel, possibly due to the production of polyols (glycerol, erythritol).

Pediococcus

These guys are Gram+, aerobic to microaerophilic spheres and are the only LAB that divide in two planes, giving rise to tetrads and large clumps. Similar to the other LAB, they have complex growth-factor requirements. Pediococci are homofermentative (convert glucose into lactate and CO₂). They are commonly found in various plants and in their respective products, like, for example, in cabbage and sauerkraut, cucumbers and pickles, grapes and wine, and wort and grain mashes. In wine, their survival is favored by a pH greater than 3.5.

Pediococcus species are undesirable in wine because: 1) they produce acetoin and diacetyl, which can give undesirable aromas; 2) they degrade glycerol to a compound that reacts with anthocyanins and gives a bitter taint (acrolein); and 3) they produce extracellular polysaccharides (ß-glucans), which cause an increase in viscosity. An extreme case of viscosity, caused by pediococci, which are highly tolerable to alcohol, is called "ropiness."

Although growth of most Pediococcus species is undesirable in wines, certain researchers have isolated Pediococcus from wines that were not considered spoiled. For example, P. parvulus altered the bouquet of a non-MLF Cabernet Sauvignon but did not spoil it. Thus, some pediococci may add desirable flavors and aromas under certain circumstances.

Identification
of Bacteria in Wine

Most bacteria growing in wine have been identified by traditional microbiological techniques, such as plating them in a favorable nutritious medium. But as the authors point out, this can yield ambiguous results, since many bacteria have similar nutritional needs and can grow under similar conditions. As a result, current research is trying to identify the species present in wine using much more powerful techniques, which rely on DNA sequencing.

Public Health Concerns

Some LAB are able to decarboxilate a variety of aminoacids, producing the corresponding biogenic amines (histamine, tyramine, putrescine, cadaverine, phenylethylamine). Biogenic amines have been identified as compounds causing negative physiological effects in humans. The presence of high concentrations of these in wine is thought to be responsible for the headaches and flushing sometimes experienced by consumers. It has been suggested that histamine doses as low as 8 mg/L can start producing headaches; and even though levels in wine are normally lower than that, some wines can contain as much as 30 mg/L of histamine. Still, the authors explain the difficulty to establish an overall toxic threshold, which depends on the potential presence of other potentiating compounds (ethanol, aldehydes, polyamines). Certain Pediococcus species and O. oeni are able to produce histamine. High levels of histamine have also been detected in wines undergoing MLF, as well as wines in extended contact with yeast lees, probably due to the increased level of aminoacids available for decarboxylation by LAB. Even though biogenic amines are currently not regulated, the concern has already triggered the production of malolactic starter cultures that do not contain the enzyme responsible, aminoacid decarboxylase.

Another compound of public health concern is ethyl-carbamate, a weak carcinogen found in many fermented foods. Some LAB can produce precursors that, after reacting spontaneously with ethanol, give rise to ethyl carbamate. One of the most common precursors is the result of arginine degradation (citrulline). For this reason, there is some interest in the development of a malolactic starter based on a non-arginine degrading strain of O. oeni.

Interactions Between Bacteria and Other Wine Microorganisms

There exists a specific order in which different microorganisms tend to grow in a wine. Non-Saccharomyces species grow during the early stages, but rapidly decline because of lack of oxygen and high alcohol. This leaves Saccharomyces cerevisiae to complete the fermentation. At the end of fermentation, when Saccharomyces enters the stationary or death phase, Oenococcus increases and conducts malolactic fermentation. After completion of MLF, other bacteria, such as Acetobacter, Lactobacillus and Pediococcus, can take over. These stages overlap, giving rise to interactions between different types of bacteria, as well as between bacteria and yeast.

Many LAB species produce antibacterial substances (bacteriocins) able to react against closely related species. For example, the growth of certain Pediococcus inhibits O. oeni, leading to problems in inducing MLF. Because of increasing consumer resistance to SO₂, the use of bacteriocins as preservatives has generated some interest. Researchers, so far, have been able to express the gene of a bacteriocin in S. cerevisiae. This raises the interesting concept of developing yeast strains able to ward off spoilage bacteria. However, the use of genetically modified organisms in wine does not seem an appealing option to consumers right now.

Some authors have suggested that the uncontrolled growth of certain Lactobacillus may lead to sluggish fermentations. For example, rapidly-growing Lactobacillus ("ferocious lactobacillus") were able to produce enough acetic acid in two to three days to inhibit yeast metabolism. But others could measure only very low levels of acetic acid in sluggish or stuck fermentations. So even though acetic acid may participate in the inhibition of yeast, other mechanisms are likely involved.

In addition to the inhibition of yeast by bacteria, there have been reports in which it was actually the yeast that were inhibiting bacteria viability (causing O. oeni to decline from 10⁵ to undetectable soon after yeast inoculation or even when they were co-inoculated). Two theories have been proposed to explain this phenomenon. In the first theory, the rapid growth of Saccharomyces removes nutrients necessary for the malolactic bacteria, which is nutritionally fastidious. But when researchers added back nutrients, the bacterial inhibition was not relieved. So a second theory states that the yeast produce toxic metabolites for O. oeni. These metabolites could include: 1) SO₂, only produced in amounts inhibitory to MLF by certain yeast strains, 2) antibacterial proteins/peptides and 3) medium-chain fatty acids, such as decanoic acid.

At the end, the authors present a summary of the chapter. AAB (Acetobacter and Gluconobacter) can cause wine spoilage through the production of acetic acid and ethyl acetate; Lactobacillus can cause spoilage through increased VA and other off-flavors; Pediococcus causes spoilage through the production of acetoin and diacetyl, as well as polysaccharides. Some species of LAB (Lactobacillus, Oenococcus and Pediococcus) can form biogenic amines and ethyl carbamate, which are a health concern. Not all LAB are "bad guys" since O. oeni is the agent responsible for MLF. Interactions among microbes can be beneficial or detrimental for a wine. Examples of beneficial interactions are the stimulation of LAB growth by yeast lees or the inhibition of Pediococcus by Oenococcus oeni. Examples of detrimental interactions would be the inhibition of S. cerevisiae by Lactobacillus and the inhibition of Oenococcus oeni by S. cerevisiae. The one who said, "It's a jungle out there" was right. wbm

Bibiana Guerra, PhD, worked at Woodbridge by Robert Mondavi as a research winemaker and a grower/educator for seven vintages. Before that, she worked at Sonoma's Flowers Vineyards & Winery, first as an assistant winemaker (1998), then as vineyard manager (1999). She is currently a technical writer in the department of enology and viticulture at UC Davis.