Winemakers have many tools for figuring out what’s in their wine.
They get amazing mileage out of just sniffing and sipping, can be uncannily accurate in predicting degrees Brix, acidity and pH simply from chomping on grapes, and are able to spot unwanted malolactic fermentation from a mile away. Just in case, every winery performs a battery of basic lab tests, and the ones that can afford it get into plating and peering through microscopes as well as running tannin assays. This arsenal of measurement techniques supports the production of a huge amount of good to great wine every year, but it’s so 20th century—and some of it even 19th century.
If you really want to know what’s in your wine, or what’s on the leaves and berries in your vineyard, you need to get with the pyrosequencing program, analyzing the DNA of every last critter that’s in there or on there. These techniques—broadly lumped together as PCR analysis—are not yet ready to be rolled out to your garden-variety 5,000-case winery, but they are on their way and destined to move from research labs to commercial wineries in the not-too-distant future. The technological power is expanding rapidly, and the costs are dropping even faster. No lab test is the silver bullet for making perfect wine, but genetic sequencing is opening up a whole new world of possibilities for understanding what’s going on in that vineyard, that fermentor and that bottle.
DNA in the news I got a peek at 21st century lab work with a roomful of other folks at the annual RAVE (Recent Advances in Viticulture and Enology) event held at the University of California, Davis, in March. David Mills from the Department of Viticulture & Enology and Johan Leveau from the Department of Plant Pathology did a dual presentation about the advantages of the new technology, including some tantalizing snippets of initial research. Mills was an early adopter of PCR—polymerase chain reaction, a method of finding and replicating tiny fragments of DNA for analysis—in the field of wine studies. A decade ago, he began using a technique known as quantitative PCR to get definitive answers about how much of a particular life form is present in a sample of wine or juice. With quantitative PCR, a target is established—the DNA template for the thing you are looking for—and then the machinery goes out and finds it and counts it in the sample. The technique proved very useful for determining, for example, how much Brettanomyces was in a wine and led to the development of assays for yeast strains, acetic acid bacteria and so on. PCR at this level already had significant advantages over traditional plating and culturing.
Some things simply do not thrive on plates, and even those that do mean the researcher is looking at a population once removed from the original sample. Perhaps the main limitation of quantitative PCR is that wine—and most of the really interesting things in life—is much more complex, full of dozens and hundreds of microbes in a dynamic ecosystem, including things no researcher knows to look for. Using what Mills and Leveau called comprehensive PCR, it’s possible to ask the machinery—in this case, a Roche 454 sequencing system—to check out and report back on everything in the sample, all the DNA of all the critters, not only the slam-dunk usual suspects but any novel microbes that might be on the scene. This sounded like pretty cool stuff to me, and while I was mulling over how to write up the presentation for this column, an article in The New York Times in late April drove home the point. The write-up discussed the publication in the journal Nature of a study by a team from the European Molecular Biology Laboratory in Heidelberg, Germany, suggesting that the ecosystems of microbes in the human gut fall into three distinct types that are not dependent on age, location or ethnicity. The idea that human digestive systems may work in different, knowable ways has enormous implications for diet and nutrition as well as the prevention and treatment of disease. Who knows? It might even supply a scientific validation for the kernel of truth in the now-discredited ancient theory of the four humors—black bile, yellow bile, phlegm and blood—that dominated Western medicine until the 19th century. In any case, that result came through comprehensive PCR and related techniques.
“The gut folks have really figured this out, gotten their forces together and said, ‘We really want to use this technology,’” Leveau said.
A week later, as I was sitting down to write this, the Times struck again, this time reporting that scientists had published research in The New England Journal of Medicine establishing a link between armadillos and leprosy, finding the same strain of Mycobacterium leprae in several infected patients in Louisiana and in wild armadillos in the area. How did they figure this out?” Whole-genome resequencing,” yet another variation on my theme. Now if researchers using cutting-edge PCR can learn stuff this interesting and important from stool samples and cuddly armadillos, just imagine what they could do with wine. Wine, unveiled Mills and Leveau concentrate on a small section of DNA called ribosomal RNA (rRNA), a component found in the DNA of all cells in all life forms. rRNA controls the synthesis of proteins and carries critical taxonomic information, allowing researchers to know what kind of critter the DNA fragment comes from, even if they don’t know everything about the critter.
It reminds me a bit of the header bytes in every computer file; they will tell you that a particular file is a JPEG image, and that it’s 32 megabytes of stuff, but don’t reveal whether it’s a reproduction of the Mona Lisa or a picture of your neighbor’s baby. Leveau tried the technique out on the material sampled from the surface of grapevine leaves and grape berries, and not surprisingly he found a huge number of different life forms. What was not so predictable was finding, in addition to standard vineyard microbes, a large amount of novel stuff not found in any current DNA database. The rRNA gave clues about the biological family and even the genus of these strangers, but then, that only gets us back to the Mona Lisa/baby conundrum. Friendly bugs? Bad bugs? Do we really know what’s on our grapevines? Mills put the machinery to work on samples taken at various points during and shortly after a fermentation, and again he found not only standard microbes but a considerable number of things not covered in any enology text and not well described in broader literature. Maybe most intriguing, Mills’ readouts showed some interlopers that seemed to grow in numbers over time, running against the trend of most microbes to die off in the presence of rising ethanol. These are simply trial runs on limited samples and are more proof of concept than ecosystem profiles. The news here is not the identification of hitherto unexpected strains of Polysyllabicus obfuscata, but the demonstration of the power of the techniques. Leveau said that what he would like to do with this technology is track the entire collection of microbiota from leaf and berry into the cellar and in the bottle over time.
No one has come close to doing that yet; indeed, one of the few published studies using comprehensive PCR to look at any kind of fermented product focused on pearl millet slurries. Yum! PCR and pyrosequencing in all their forms have limitations, of course. One important gap, according to Mills, is that “none of these methods differentiate between live and dead cells.
There is no published, peer-reviewed evidence that standard quantitative PCR can distinguish live and dead.” As you might imagine, people are working on this. Another limitation is that comprehensive PCR only gives the relative numbers of different microbes, not the actual, absolute counts, though that can often be derived from going back and using quantitative PCR to search for a specific critter. Leveau notes that sample preparation is critical to ensure, for example, that the analysis is performed on things sitting on leaves and berries, not things inside them. At this point, there are no universal protocols for sample preparation, which gets done case by case. The ability to interpret results from a PCR run depends on the quality of the reference databases used to match the findings.
With the explosion of technology and the multiplication of the teams of researchers using it, Leveau says dealing with the huge amount of data is a struggle in itself. And one more problem: While these methods have gotten dramatically cheaper, they are not free. Putting one of these high-powered machines to work on a set of samples for an eight-hour sequencing search costs several thousand dollars. Systems like the Roche 454 are quite expensive; Davis, for example, doesn’t have one, and relies on a centralized analytical services provider to do the actual runs. Tapping the potential of these techniques requires major financial resources, and it would behoove the wine industry to make sure Leveau, Mills and their kin spend their time searching for microbes, not for donations.