Indigenous yeast in Sauternes, multi-species family wineries, and Casale’s Trebbiano orange wine

A group of French scientists, mostly based in Bordeaux, have recently published evidence* that the same Saccharomyces cerevisiae populations have lived at their winemaking homes in Sauternes for at least 23 years.** Two decades is a brief moment, whether you’re measuring in human-history-of-Sauternes time or in yeast-evolution time, but their work still supports the idea that yeast populations become associated with wineries and stick to them.

A main point the scientists aim to make is that these longstanding, tradition-following Sauternes estates haven’t been badly polluted by modern, commercial winemaking yeasts. Only seven percent of the S. cerevisiae strains they collected could be clearly connected to specific commercially produced strains. This news is excellent for those in the natural wine camp who want to call spontaneous ferments “natural” or “wild.” In this little corner of Sauternes, at least, it seems that wines allowed to ferment spontaneously aren’t just being fermented by commercial yeasts persistent in the environment. Most of the work of fermentation is indeed being done by yeasts which very likely existed in the area prior to Lallemand and the rest introducing their tidy little foil packets of active dried specimens.

A good case can be made, I’d argue, for calling those non-commercial yeast something like “domesticated indigenous” rather than “wild:” human winemakers have selected and bred them up for desirable winemaking characteristics over many generations, just as human dog-keepers have bred up select canine features or orchid aficionados have carefully developed new plants. Domestication can happen even if living things are never bought and sold.

Regardless of what we choose to call these yeast species, though, this sort of research says that those species are part of the traditional winemaking environment, part of the terroir. Terroir is apt because that complex term incorporates human traditions and activities, the way winemakers and the rest of the community have shaped the land and its various characteristics: soil, plant life, microbial life, architecture, machinery, maybe some canines. Terroir doesn’t force us to make a choice between those winery-associated yeasts being wild or cultivated. They’re both, and winemakers and winemaking have evolved with them.

The Casale orange trebbiano*** I opened last night is made by a winery with whose winemaking history in their neck of Chianti can be traced back to 1770. The wine itself is a 2012 vintage and the current release, kept on the skins for 30 days and in stainless steel for two years. The property is biodynamic, the wine spontaneously fermented.

Even if the mid-palate is fairly empty, the nose and the finish are more than enough to rescue that deficiency: savory, nutty, sweetly orange blossomy and honeyed up front all at once, and acidity with great energy and tension (minerality, if you wish) on the tail, with a pleasant light astrigency across the breadth of the middle. This isn’t just a wine for thinking with, and I’d readily pour it for anyone who thinks that all biodynamic orange wines must smell like hair salons and taste like dirt, because it doesn’t. But it’s a beautiful example of a wine good both for drinking and for thinking, that tastes pleasant and nourishing and that provokes pleasant and nourishing trains of thought.

When we care about family wineries – for the purely aesthetic value of tradition as well as the economic value of maintaining local businesses and the winemaking value of passing down knowledge and physical infrastructure – we should care about the extended family, not just the humans but the yeast (and maybe even other species) who have also lived on a winemaking property for generations, who make the wine with their human coworkers. Taking care of the family can’t be just about loving your brothers, or even about supporting the other families who might work with you, but about caring for your non-human brothers and coworkers. Now, that’s a pleasant and nourishing train of thought.

*Full text of the article behind an academic journal’s paywall, unfortunately.

**A bit more detail on that point: the team collected samples from three Sauternes estates over 2012, 2013, and 2014, isolating 653 individual yeast strains, and compared them to commercial strains and to 102 “library” isolates collected from one of the three estates between 4 and 23 years ago. The comparisons to determine which strains are related to which other strains rely on 15 “microsatellites,” or specific, small sections of DNA the sequence of which is highly and characteristically variable amongst strains. Much could be said about how we define what constitutes an individual “strain” on the basis of the tools we have available to do so – appearance (phenotype, in biology jargon) in the past, genetic today – but I’ll forgo that conversation for today.

***Purchased for £15.50 from Wood Winters in Edinburgh.

Why do yeast make alcohol?

Ever wonder why yeast make alcohol? Probably not, I realize, but you should. Yeast throw off ethanol in the process of metabolizing sugar, so alcohol is a byproduct of survival; fair enough. But alcoholic fermentation is, in fact, a surprisingly inefficient way to get energy. The standard oxygen-requiring way of breaking down sugar used by most cells, our own included, wrings somewhere between 30 and 38 ATP (38 is the ideal number; it’s probably never quite that high in practice) out of a single glucose molecule. (ATP is the cellular currency in which energy is transferred and spent.) Nevertheless, alcoholic fermentation has the distinct advantage of not needing oxygen and so it makes perfectly good, intuitive sense for Saccharomyces cerevisiae to use it when oxygen isn’t available.

Here’s the quirk: S. cerevisiae uses inefficient alcoholic fermentation even when it does have access to oxygen, even though it has the machinery for the much, much more energetically worthwhile aerobic metabolic process. Yeast will only switch to aerobic metabolism when the amount of sugar available for them to eat is very low. Why? A good question, and one microbiologists haven’t had much success answering.

Our best hypothesis according to a brand-new review on the subject comes in two parts:

  1. Alcoholic fermentation lets yeast act fast to use up the “public goods” while squirreling away private resources for later. Every microorganism you’ll encounter in grape juice can consume sugar. Very few can also consume (and get energy out of) ethanol, but yeast can. So, by converting sugar to ethanol, S. cerevisiae can starve out other microbes and leave itself with a food source for later.
  2. As an additional and maybe even bigger benefit, ethanol is toxic to most yeast and bacteria at concentrations that Saccharomyces can tolerate with relative ease

Possibly the most bizarre thing? We don’t know much about what determines the circumstances under which S. cerevisiae, our long-time compatriot and coworker, produces alcohol versus making energy in some other way. We’ve looked at when and where different yeast genes are expressed and when and where it makes different byproducts but, like so much else in the wonderful and frustrating world of modern-day genetics, putting together the whole story is still a work-in-progress.

S. cerevisiae: Friendly to us, but…

Saccharomyces cerevisiae is one of the friendliest microbes around. The good, old, familiar yeast used for (nearly all) bread, ale (lager yeast are different), (nearly all) wine, saké’s second (alcoholic) fermentation, and an improbable array of industrial applications: making CO2 to bubble into aquariums and keep subaquatic plants from suffocating and manufacturing insulin, for example. Heck, my favorite state even elected it as the official state microbe this past May, making Oregon the first – and only – state to have a patron yeast.**

The yeast so friendly to us, though, is a good deal more sinister to its peers.

Branco and colleagues, at the Laboratorório Nacional de Energia e Geologia in Portugal, have figured out that some S. cerevisiae strains release peptides that kill other types of yeast and bacteria found in the early stages of wine fermentations, including the love-it-or-hate-it Brettanomyces bruxellensis (responsible for barnyardy Burgundy and a whole lot of spoiled wine that tastes like old gym socks).

Microbiologists have also known for decades that some strains of S. cerevisiae are “killer yeast,” releasing peptides (very small proteins) that cause the death of other “killer sensitive” strains. Killer yeasts can be a cause of stuck fermentations: if your carefully selected yeast is killer sensitive, an accidental contaminant population of killer yeast could wipe it out, leaving only the less well-adapted accidental yeast unable to finish fermentation, and leaving residual sugar in a wine that was supposed to be dry. Brewers who want to play with wine yeast in ale recipes need to be careful, too, because the majority of wine yeasts release killer peptides and all ale yeasts are killer sensitive; mixing the two requires choosing a non-killer wine yeast. More recently, we’ve learned that some malolactic fermentation bacteria (like Oenococcus oeni) are also sensitive; a reason, perhaps, for some malolactic fermentations not kicking off as expected.

So we knew that already. But the new peptides that Branco and colleagues have identified are important for two reasons.

First, we’ve been operating under the general assumption that killer yeast will only kill other yeast (and maybe some of those malolactic bacteria). If that’s not true, we may have a new and improved explanation for why some microbes interact the way they do.

Second, if some S. cerevisiae make a peptide that will kill off undesirable or spoilage yeast, we could use it to our advantage. Produced in bulk, they could be added as a “natural” preservative against Brett and other nasties. Other yeast toxins have been used in similar ways.

S. cerevisiae always takes over wine fermentations, even when musts aren’t inoculated with commercial active dry yeast, and even though grapes in the vineyard play host to a large and diverse population of other yeasts and bacteria. We’ve generally assumed that S. cerevisiae takes over because it’s best able to withstand the rather noxious fermenting grape environment – high sugar, high acid, increasing amounts of alcohol (S. cerevisiae‘s high alcohol tolerance is especially significant) – and because it grows fast.

What this research calls to my mind are recent explorations into adding microorganisms other than S. cerevisiae to fermentations, not to carry the burden of sugar-to-yeast conversion, but to add flavor and complexity. New killer peptides may offer a welcome explanation for why some microbes work well with S. cerevisiae and others don’t. For at least some of those that don’t make it, maybe we need to change the cause of death from “suicide under high stress” to “murder.”


**Wisconsin made an attempt in 2010 to elect Lactococcis lactis, the workhorse bacteria for culturing cheddar and many other cheeses (not to mention lactic-fermented pickles, but those are more Seattle than Wisconsin), but the movement stalled in the Senate after being passed by the House. And Hawaii’s House made a movement to elect Flavobacterium akiainvivens (which doesn’t appear to be of any particular good to anyone, but was found on a rotting bush on O’ahu) in February but, again, it failed to clear the Senate. The official state microbe is evidently catching on, to which I give an earnest “Hooray!” if it means more popular awareness for how much bacteria and yeast contribute to our livelihoods. The marvelous Elio Schaechter thoughtfully collected crowd-sourced suggestions for some other states on the dedicated microbe blog Small Things Considered here