New, better pictures of what bacteria are doing during fermentation

Short: Microbiologists, using new techniques, are finding that actively fermenting grape musts contain a much wider variety of bacteria than we’ve previously recognized.

Longer: Yeast, being the main actors in alcoholic fermentation, are going to get most of the attention during it. But that’s not the only reason why we don’t know a whole lot about bacteria during fermentation. They’re hard to grow. Doing microbiology the classic way means collecting samples and growing bacteria from them in petri dishes, then identifying whatever grows in the dish. I used to spend measurable fractions of my life “making plates” by mixing Jell-o for bacteria and pouring it into hundreds of little plastic dishes. A main activity of microbiology is waiting for stuff to grow on your plates. The logic here is simple, understandable, and incredibly silly. Bacteria are unbelievably diverse and incredibly specific to their environments. It’s balderdash to think that all of them are going to grow happily, on command, in a dish, in a week or two (if you don’t just throw Monday’s plates out on Friday, which sometimes happens).

The petri dish method is just fine* for working with well-identified and properly house-trained bacteria. It’s pretty horrible for trying to identify all of the unknown bacteria growing in some mystery environment. Only when alternatives became available did microbiology really start coming to terms with the magnitude of what it had been missing. Today, looking for bacterial “unknowns” means identifying bacterial DNA, which is more direct and gives you a better chance of picking up punk microbes that aren’t willing to grow nicely in captivity. Search for bacterial DNA in a vat of actively fermenting grape and you’ll find evidence of a lot more bacteria than the conventional mechanism ever had us thinking about.

Using these techniques, Bokulich, Mills, & co. at UC Davis have been mapping bacterial communities in wineries around the calendar year, wineries across California, and wines with more and less SO2. New research (open-access article), from a (mostly) Washington state-based group, has pointed out something simpler and yet very worth knowing: fermenting wine contains scads more bacteria then we’d ever thought about before. They used “next generation sequencing”** techniques to take snapshots of bacterial communities five times through two weeks of fermentation.

The authors make some questionable comparisons of patterns of bacterial growth between their two study conditions – all the grapes involved were organically grown Riesling, but half were fermented “organically” without added SO2 and the other “conventionally” with SO2. But the experiment involved only a single comparison: two vats, same batch of grapes. Limited replications is no doubt a trade-off with fermenting in realistic 15,000 gallon volumes instead of the completely unrealistic five-gallon carboys too common in much wine research. Regardless, it’s going to take many more comparisons before it’s possible to talk meaningfully about differences in bacterial abundance with and without added SO2.

Here’s why this research is still important. Right now, wine bacteriology is mostly two things: malolactic fermentation, where bacteria are the good guys (unless you’re trying to prevent MLF and it’s happening anyway), and spoilage by a pretty well-known set of culprits, especially acetic acid bacteria. That’s a bit like saying that all Americans are either New York City firefighters or drug dealers. There is a whole lot more going on in both cases. And some of our persistent wine mysteries – why some fermentations stick, some go faster or slower, some produce one aroma and others another – may owe something to that unseen majority. If microbiologists start seeing them, maybe we’ll find out what.

*It’s also time-consuming, labor-intensive, and incredibly wasteful in terms of the masses of plastic that get thrown away. Sometimes you want to do an experiment and can’t because you don’t have enough plates. Or the results you get at 6:00 pm suggest an experiment you should do tomorrow for which you’ll need more plates and you stay until 9:00 pm going through the several-hour process of making more, or someone else uses your plates without telling you, or your plates get contaminated with mold and you have to throw a big batch out.

**As opposed to “deep space nine” sequencing techniques, which are expected to come out next season, will take longer and be more sophisticated, but will never be quite as cool as its predecessor because Patrick Stewart isn’t involved.

Brett + bacteria = worse, or better

Microbiology has gotten a lot wrong studying yeast and bacteria. We’ve assumed, until quite recently, that if a microbe doesn’t grow in a dish it’s not there. And that a microbe is either on/live/growing or off/dead. And that we can study microbes in isolation — “pure culture” — away from other species in little sterile dishes and expect them to behave normally. In all fairness, microbiologists have sometimes seen these as a problems, but have mostly just gone on this way, writing books about what we think we know.

DNA detection and sequencing technology is showing just how many bugs don’t grow in dishes — “high throughput” technology can document (theoretically) all of the species in a drop of [insert favorite liquid here]. That’s pretty routine these days. And we’re slowly beginning to study how mixtures of microbes — you know, the way they live in the wild — behave in the lab. Wine was a bit ahead of the curve here: microbial enologists have been studying the goings-on of spontaneous and mixed fermentations since the late 1980’s.*

Continue reading

Studying sulfur dioxide effects with better DNA technology suggests we may not need much of it

Fast: In a new study using better-than-ever microbiology, 25 mg/L SO2 added after pressing was enough to “stabilize” yeast and bacterial growth during fermentation, and higher concentrations actually seemed to slow fermentation. Inoculating the must with commercial S. cerevisiae had a very similar effect, even without adding SO2, which looks really, really good for no-added-sulfur wines. BUT (and this is a big but) the study only included one wine (a California chardonnay) made in one way, in smallish (19 L) lab volumes. Goodness only knows if their results will generalize, but let’s hope this encourages someone to look.

More: Sulfur dioxide is the single most commonly used winemaking chemical worldwide. That familiarity probably has something to do with our not understanding it better: we know it’s safe, we know how to use it, and so we don’t have much reason to study it.

In all fairness we do understand SO2 well, but microbiology keeps changing. The publishing dynamo* of Nicholas Bokulich and David Mills – responsible for really excellent recent research on how microbes are spread around a working winery over space and time – plus UC Davis wine microbiologist Linda Bisson (and another Davis student and a Japanese collaborator have published a new American Journal of Enology and Viticulture article on how SO2 affects bacteria and yeast populations in fermenting wine.

The question isn’t new, but the technology they’re using is. Short story: better DNA detection techniques let them pick up on the presence of a bigger range of both bacteria and yeast than previous strategies.

Longer story: Microbes in wine (and elsewhere, for that matter) can be “viable but nonculturable” (VBNC), a new idea ten years ago when microbiologists could still think that agar and Petri dishes were a reasonable way of identifying bugs in a sample. Until better DNA technology made clear a serious issue: yeast and bacteria might be stressed out enough by environments (like wine) to not grow on command but still be alive and able to multiply and cause problems, aka VBNC. (The unculturables who won’t grow in dishes at all are trouble, too.) The details of the high-throughput DNA sequencing they used to ensure that VBNC bugs weren’t left out of their survey aren’t important except to note that it lets them detect more microbes than previous studies.

The other great = new element of this study is its looking at multiple SO2 concentrations, from 150 mg/L down to nil. More work for them; more data for everyone. They also included ferments inoculated with commercial S. cerevisiae and not, which ended up being important.

Their results say that the most important factor in determining what grows in fermenting wine seems to be the degree to which a single strain has the opportunity to take over. One way of encouraging dominance is inoculating with commercial yeast: it more or less takes over and overall microbe diversity declines. But another way is adding SO2, which knocks down some microbes and gives tolerant ones (S. cerevisiae strains included) an opening. Adding SO2 and inoculating S. cerevisiae even without SO2 had similar effects on overall microbial diversity. And, moreover, 25 mg/L was enough SO2 to “stabilize” the ferment. In other words, sulfur-free wines may be less risky than winemakers are generally inclined to believe if they inoculate (which plenty of people inclined not to use sulfur are also inclined to avoid).

The obvious problem: one wine, one vintage, one set of processing techniques, and 19 L volumes. All of these are major reasons to question whether these results will hold for any other set of circumstances. pH and a slew of sulfur-binding compounds affect SO2 efficacy. Fermentation temperature, oxygen, clarification, means of harvesting…the list of processing steps important to microbial diversity is too long to list. And it’s well-known that fermentation volume is important to microbial kinetics.

In short? This article is almost certainly more important to wine microbiologists as a methods paper than to winemakers. (It’s not incidental that the methods section abbreviates the winemaking protocol — “grapes were harvested, crushed, and pressed according to standard winemaking procedures,” whatever that means – and uses nearly a full page of text to describe the DNA sequencing technique.) Nevertheless, it may well serve as impetus for more experimentation with low- and no-sulfur wines, and a good reminder that we always have more to learn about SO2.

Even more: find the full paper, with many more details on which specific yeast and bacteria species were detected and when they peaked (unfortunately behind the AJEV‘s lovable paywall) here. Read the full paper if you can; it contains plenty of potentially idea-generating details that I’ve not even attempted to summarize here.

*Hackneyed maybe, but, seriously, what else do you call them? Bokulich’s CV, as a PhD student, could put to shame plenty of tenured professors. When I’m not just feeling horribly inadequate, I’m wondering where this guy will end up post-graduation. Barring his speaking French and fancying living overseas – or starting a lucrative consulting firm – he’s probably lined up to make tenure at Davis in record time. Heck, he probably already qualifies for tenure.