Subscriber login Close [x]
remember me
You are not logged in.

Yeast - wild, cultured, genetically modified - by Jamie Goode

Published:  23 July, 2008

A few months ago, I visited one of Australia's leading small wineries at harvest time. A small open-top fermenter had been loaded with Shiraz the previous day: now, some 24 hours later, the must was already quite deeply coloured, but fermentation hadn't yet begun.

I tasted some of the pink/purple juice. It was very sweet, quite simple and a bit like concentrated grape juice, which is, I suppose, what it

was - but my point here is that it bore very little relationship to the complex, thought-provoking (and expensive) liquid it would become by the time it was bottled. The primary agent of this transformation? Yeasts. Unicellular, microscopic fungi that are responsible for alcoholic fermentation. As well as the transformation of sugar into alcohol and carbon dioxide, the complex chemistry occurring during this fermentation process also involves the formation of myriad flavour molecules from precursors present in the must. Figures I've seen quoted suggest that of the estimated 1,000 or so volatile flavour compounds found in wine, at least 400 are produced by yeasts. Without yeasts we wouldn't have wine.

From the earliest days of winemaking, some 8,000 years ago, until the middle of the 19th century, this fermentation process must have seemed almost magical. It was only 150 years ago that Louis Pasteur identified yeasts as the agents of this change. And it wasn't until the 1970s that cultured yeasts were routinely used in winemaking.

Let me pause here to make a rather indulgent (but perhaps important) semi-philosophical point. Considering the importance of yeasts in moulding the flavour of wine, it's remarkable how little attention they get. Vines and grapes are easy for us to understand because we can see them, and they get all the attention. But if yeasts were bigger - big enough that we could see them - we'd read a lot more about them. But yeasts, along with other microbes, are much harder for us visually dominated humans to get a handle on because of their microscopic size. If we want to have a proper perspective on winemaking we need to give more attention to these microbes. So perhaps we should imagine that fermentation vats are the size of a small ocean, and yeast cells are the size of a modest potato, but capable of growing rapidly and dividing every half hour or so. Suddenly they don't seem so abstract.

What occurs during a 'natural' fermentation?

Let's stretch our potato analogy to try to illustrate the fermentation process. Imagine that the composition of the ocean can vary, and that rather than one type of 'potato', we have literally dozens, each type with subtly different growing abilities (and conveniently colour coded). Initially, the ocean has just a few of each type of potato bobbing around, with the occasional potato plopping in from the sky. Its surface is largely calm. Then, within 24 hours, these potatoes begin to grow and divide fast, such that within a short time the ocean is bubbling full of rapidly multiplying potatoes, feeding off the contents of the ocean and in turn releasing their own waste products (the word 'fermentation' derives from the Latin verb 'to boil'). There may be as many as 5 million yeast cells in each drop of fermenting grape juice.

At first, it's the red, green and yellow potatoes that thrive in this seething ferment, but within a few days blue potatoes start appearing. The non-blues die and fall to the bottom, and the blues take over, continuing their individual cycles of growth, multiplication and then death. After a couple of weeks things start to calm down, and then the potatoes are almost all gone - they've sunk to the bottom of the ocean and formed a thick carpet of dead and dying spuds, which continue to exude their contents into the ocean. However, a number of potatoes still remain floating around in a semi-dormant state - these are of a number of different colours. It's as if they are waiting for things to change; hanging around in case there's some action.

Like the potatoes of this analogy, yeasts and bacteria are ever-present in the winery environment. Even in what appears to be a spotlessly clean cellar there will always be some receptive surface, such as an uneven soldered joint in metal pipework, where microbes can hide. Barrels are a particularly happy niche for fugitive microbes because the structure of wood means that, in practice, it is almost impossible to sterilise it. Because a potential source of inoculation is just about ubiquitous, all yeasts and bacteria need are the right conditions, and they will begin to grow. Grape must represents a sugar- and nutrient-rich medium that's ideal for the growth of certain microbes. As it ferments, it changes, and its suitability for one species or strain wanes as its suitability for another develops. This is an important concept: create the right conditions and you can select for the population of organisms that you want to be growing at that particular time. Winemakers tend to concentrate on eradicating rogue organisms from the winery. This is a good idea, but at the same they have to make sure that the musts they are working with give a competitive advantage to the sorts of bugs they'd like to see growing.

Still, the issue of cleanliness is important. It's a numbers game: the size of the inoculum (how many yeasts and bacteria are present initially) will be an important factor in determining whether an infection will be a problem or not. It matters who gets in first - whether it's the microbes you want to be growing there, or those that you don't. Related to this is the issue of ensuring a healthy ferment. Yeasts require a nitrogen source and adequate oxygen, as well as a range of micronutrients. If these aren't present then the result is a struggling or stuck ferment, which can later cause 'reduction' problems in the wine.

Yeasts are also widespread in the vineyard. They spend winter in the upper layers of the soil, spreading to the vines during the growing season via aerial transmission and insect transfer. They colonise grape skins during the maturation phase, although they never reach very high levels on intact grapes. Contrary to popular opinion, the bloom on the surface of grape skins isn't made up of yeast populations, rather a wax-like scaly material that doesn't harbour many fungi.

Only a limited number of yeast species are present on grapes, the so-called 'native' yeast populations. These include: Rhodotorula, the apiculate yeasts Kloeckera apiculata and its soporiferous form Hanseniaspora uvarum (the most common by far), and lesser amounts of Metschnikowia pulcherrima, Candida famata, Candida stellata, Pichia membranefaciens, Pichia fermentans, Hansenula anomola. Also present may be potential spoilage organisms such as Brettanomyces. It needs to be added that yeast nomenclature is rather a confusing business, with various names used for the same bug. This isn't surprising, because until the recent development of molecular methods for typing yeast strains and species, it was rather hard to tell all of them apart. Unlike the potatoes in the analogy, yeasts aren't neatly colour coded.

One interesting observation is that the main wine yeast, the alcohol-tolerant Saccharomyces cervisiae, is rare in nature. Attempts to culture it

from the skins of grapes have proved unsuccessful. The only way its presence can be demonstrated is by taking grape samples and placing them in sterile bags, crushing them under aseptic conditions and seeing what happens, an experiment that has been done in Bordeaux. At mid-fermentation, S. cerevisiae, which is undetectable on grape skins, represents almost all the yeasts isolated. In a few cases no S. cerevisiae is present and apiculate yeasts do the fermentation. Opinion still remains divided about whether S. cerevisiae originated in the vineyard or is actually a human introduction. More on this later.

Wild yeast populations

The early stages of 'wild' (uninoculated) fermentations is typically dominated by Kloeckera, Hanseniaspora and Candida. As the alcohol levels rise a little, these players bow out and others such as Cryptococcus, Kluyveromyces, Metcschnikowia and Pichia step in to take their turn. It has been estimated that in an uninoculated ferment as many as 20-30 strains participate. But as alcohol levels reach 4-6% the native species can't cope with the hostile conditions, and the alcohol-tolerant S. cerevisiae will take things onwards from here. So the key difference between natural ferments and those carried out by cultured yeast inoculations is in the early stages of fermentation.

There are complications, however. Most winemakers add some sulfur dioxide (SO2) on crushing, to reduce the risk of oxidation and also to kill off rogue microbes. This will slant things in favour of S. cerevisiae and the more robust of the native species, eliminating some of the less desirable wild yeasts and spoilage bacteria, which tend to be more sensitive to the microbicidal actions of SO2. Temperature also affects the balance of yeast species in the fermentation. Cooler temperatures (below 14C) favour wild yeasts such as Kloeckera, whereas higher temperatures shift things in favour of S. cerevisiae.

Aside from the properties of the wild yeasts themselves, spontaneous ferments cause a delay in the onset of vigorous fermentation. In red wines this will allow oxygen to react with anthocyanins and other phenolics present in the must, enhancing colour stability and accelerating phenolic polymerisation. This enhanced exposure to oxygen before vigorous fermentation is underway could have important flavour and textural effects on the wine thus produced, quite separately from the characteristics contributed by the diverse set of wild yeasts carrying out the fermentation.

Wild versus cultured yeasts

There's a bit of an ideological divide between those winemakers who advocate natural fermentations and those who choose to use cultured yeasts, although many winemakers sit somewhere in the middle. At one pole we have the likes of biodynamic guru Nicolas Joly, who maintains that 're-yeasting is absurd'. At the other, there are many who share the views of yeast researcher Sakkie Pretorius (currently director of the Australian Wine Research Institute, AWRI), who describes the risks involved with wild yeast ferments as 'staggering'. Estimates are that worldwide, around 80% of fermentations are natural, and 20% with cultured yeasts, with the latter gaining ground. This pattern loosely follows an Old World-New World divide, with the former largely preferring to use indigenous yeasts and the latter relying on cultured strains, although this split is far from an absolute.

Let's examine the Pretorius position first. There are three arguments against allowing natural fermentations. First, risk. Because the initial inoculum of yeasts from the winery environment and grape skins is quite low, fermentation can take a while to get going. This introduces an element of risk: if bugs such as Acetobacter (the acetic acid bacterium that turns wine to vinegar) establish themselves before the fermentative yeast species, then the wine will be at risk of spoilage. Also, there's no guarantee that the native yeasts that establish themselves will do a good job. Like S. cerevisiae, all the various native yeasts exist in many different strains, some desirable, others not. With a spontaneous fermentation you take what you are given.

The second objection is that some winemakers point out that very few 'wild' fermentations are actually wild. As harvest gets underway, the winery equipment will be a ready source of inoculum, and fermentations will get going a lot faster, with S. cerevisiae establishing itself sooner. Studies have shown that after a few days of harvest operations, half of the yeasts isolated from the first pumping over of a spontaneously fermented red grape tank are S. cerevisiae.

The third objection is that winemakers who chose spontaneous ferments lose an important element of control over wine style and quality. Cultured yeast strains allow winemakers to control the fermentation process, removing the risks of high volatile acidity or funky flavours that can come with wild ferments. If you are in charge of large quantities of someone else's wine, it's understandable that you'd want to play it safe. Cultured strains with specific properties can also be chosen to complement the wine style being made: for example, specific cultured yeast strains are an important contributor to the highly successful New Zealand style of Sauvignon Blanc.

So, to the Joly position. Why take the risk of a spontaneous wine fermentation? In many cases the motivation will be ideological. For many winegrowers, natural yeasts are seen as part of the terroir. 'Natural yeast is marked by all the subtleties of the year. If you have been dumb enough to kill your yeast you have lost something from that year,' says Nicolas Joly. For those who see naturalness as an important property of wine - one that separates it from 'manufactured' drinks such as beer and spirits - the use of wild yeasts is a fundamental principle.

Others do it for quality reasons: native yeasts are thought to produce wines with a fuller, rounder palate structure, and the ferments tend to be slower and cooler, burning off fewer aromatics. Wines made with wild ferments are often richer-textured and have more complexity. There is also a cost saving: cultured yeast has to be paid for.

Some of the risk associated with the slow start of natural ferments can be offset by getting a starter culture (known in France as a 'pied de cuve'), a few days before the first main harvest day. This can then be used to kick-start fermentation. However, if the starter culture has been going for too long, then the dominant yeast will be S. cerevisiae and some of the benefit of the diverse wild yeasts that only thrive at low alcohol levels found early on in fermentation will be lost.

But are winemaking yeasts all 'artificial'?

Here's a provocative thought, recently suggested by South African researcher Florian Bauer at a symposium on Yeast's Contribution to the Sensory Profile of Wine held in La Rioja, Spain (27-28 April 2005). Could it be that all fermentations are artificial, in the sense that the alcogenous yeast S. cerevisiae isn't naturally occurring, but has evolved to fit an artificial niche created by the practice of winemaking itself? From archaeological records we know that winemaking is an ancient biotechnology dating back 8,000 years, with the wine yeast the first domesticated microorganism. The activities of these early winemakers provided an artificial, rather hostile ecological niche, by creating an environment high in sugar and acid which then transitioned to one of lower sugar levels, low oxygen and relatively high alcohol levels. This winemaking environment would have provided a significant selective pressure that could have led to evolutionary adaptations that resulted in today's wine yeast.

As we've already noted, very few commercially used S. cerevisiae wine strains can be found isolated in vineyards, even where these vineyards are close to wineries that are stacked full of yeasts. This suggests that yeast strains are not well adapted to other 'natural' environments. The implications? S. cerevisiae could be a domesticated, non-natural yeast that evolved in conjunction with human activity. This conclusion is supported by the work of Alsace-based researcher Jean-Luc Legras and colleagues. He analysed 12 sections of DNA known as microsatellite loci from 400 strains of S. cerevisiae wine yeast isolated from around the globe, finding evidence for a common origin for these yeasts. The conclusion? S. cerevisiae has likely spread around the world along with the spread of the grapevine Vitis vinifera.

A third way

Is it possible to argue for a middle ground between those who see cultured yeasts as a vital winemaking tool and those who see wild yeasts as part of the terroir? I think so. In this context, the term 'terroir' might be unhelpful, because it carries with it supplementary meaning that it's the hand of nature and not the hand of the winemaker that is responsible for wine quality. It's an issue of agency. Many advocates of terroir deliberately downplay the role of the winemaker, to the point of being in denial that humans have anything useful to contribute to wine quality at all. Perhaps a more useful term here might be typicity. There's an implicit honesty about this term because it preserves the all-important concept of 'sense of place', while acknowledging that the winemaker is an important partner in the process of realising the potential of the grapes that are brought into the cellar. I'd argue that proponents of 'typicity' can, in fact, do more as guardians of the wonderful diversity of wine styles seen around the world that have as their origin local characteristics. This is because they are not constrained by an ideology that refuses to acknowledge that their decisions in the cellar can do anything useful to bring out the sense of place in a wine. Subscribing to typicity allows winemakers to use cultured yeasts with the aim of enhancing, or perhaps less controversially, bringing out the typicity - for example, making a Puligny Montrachet that tastes more like a Puligny.

A storm brewing: genetically modified yeast strains

To finish, I'd like to touch on possibly the most controversial issue surrounding wine yeasts - the use of genetically modified (GM) yeasts. Plenty of these already exist, but they haven't previously been commercialised because of the negative reactions of consumers to

GM food products, which don't seem to be budging much.

Now, however, a GM yeast strain called ML01 has been commercialised and is authorised for use in the US. This yeast, made by Springer Oenologie, has two extra genes (a malate transporter gene from Schizosaccharomyces pombe [another yeast] and the malolactic enzyme gene from Oenococcus oeni [a bacteria]), which allows it to carry out malolactic fermentation (normally done by bacteria after alcoholic fermentation has completed) at the same time as alcoholic fermentation. There are several advantages to this, one of which is that the resulting wine is less likely to contain biogenic amines which are produced by the bacterial malolactic fermentation. In the US, yeasts are classified as processing agents, and thus wines made with this yeast would need no declaration that they contained GM ingredients. In many other countries, such as New Zealand and Australia, the regulations are more stringent. Is anyone making wine using this GM yeast? If they are, they aren't telling anyone, for understandable reasons. In response to this, the AWRI has issued a statement declaring that no GM yeasts will be used in Australian wine for the foreseeable future. But because it is so much easier to produce yeasts with desirable properties by GM technology (and there are some traits that are impossible to select for by conventional breeding), research continues apace globally on GM yeast technology. It all seems at odds with wine's image as a natural product, but supporters of the technology argue that what they are doing is not creating fake wines, but unlocking the latent flavour and aroma potential of grape must by using yeasts with special properties. One yeast researcher has even gone on record as stating that the best wines are yet to come. As a scientist who cares a great deal about the future of wine, I favour a cautious approach: if GM yeasts become widespread, the danger is that wine will be seen as just another manufactured beverage. If we kill the 'naturalness' of wine, we run the risk of destroying the whole venture.