October 2010 Issue of Wines & Vines

Finding Brett Before It Finds You

Early detection is critical while preventive measures are being developed

by William Edinger
agar plate Brettanomyces
Figure 1: The agar plate at left contains an acid indicator with no growth. Growth of Brettanomyces causes color change (right).
One of the most vexing developments for winemakers in recent years has been the increasing incidence of spoilage by Brettanomyces yeasts, or “Brett.” The emergence of this age-old contaminant is undoubtedly due to a shift in current production factors that immediately preceded it: longer and longer hang times, extended maceration before and after fermentation, extended aging in barrels and a desire to reduce SO2 levels to an absolute minimum, all of which contribute to an environment where Brett can survive, grow and multiply rapidly.

The presence of Brett in a wine is frequently identified because of several notorious phenolic odors, along with alterations in acidity, color and balance. While some winemakers may consider a few aspects of Brett growth to be beneficial to complexity, and have even toyed with intentionally dosing from reserved barrels of “good Brett” wine, the great majority of the time its effect is simply spoilage.

While prevention is fairly easy—minimal time on the vine, faster processing, more tank time and less barrel time, higher concentrations of SO2, etc.—it runs counter to the stylistic trends that characterize many upscale wines. Consequently, the approach to managing Brett seems to be early detection, before the contamination has time to become a catastrophe.

However, detection of Brett contamination is problematic for a number of reasons. While its immediate origin may be from cellar equipment, especially used barrels, its ultimate origin is undoubtedly the vineyard.1 But, like S. cerevisiae, the numbers of Brettanomyces yeasts are generally so low that they are overwhelmed by other wild yeasts and hard to isolate.

Similarly, in musts and wines Brett yeasts may initially be in such low numbers that they are either undetected or not regarded as significant. Brett’s tolerance to alcohol is so high, however (even higher than that of S. cerevisiae), that after alcoholic fermentation, the environment is highly selective for their dominance, especially if sulfite is low. Extended barrel aging simply expands this window of vulnerability, and growth, with concomitant spoilage, can be quite rapid.

Biotechnology to the rescue?
Plating on agar has been the traditional method for finding and identifying spoilage yeasts, and the method has both advantages and disadvantages. Today, however, biotechnology has brought us several new approaches. In the late 1990s, a method was presented commercially that used Peptide Nucleic Acid (PNA) probes.2 This looked promising, but it required having a fluorescence microscope. A less expensive version turned out to be too difficult to interpret, and that method failed in the market. Among other approaches, RNA or DNA analyses are used in research environments, but they are not really practical for most (if any) winery labs and are probably too time-consuming to be of use in any event.

Polymerase Chain Reaction (PCR) analysis has been shown to be feasible,3 but for various reasons it has not been free of controversy. However, there seems to be little agreement about why some people have had problems with it. According to Dr. Rich DeScenzo of ETS Laboratories, the No. 1 cause of unreliable or inconsistent results in his experience has been due to the sampling methods used to collect the wine for analysis, which is true of all micro-testing.

The main problem with PCR is that it only detects the target organisms that primers or primer/probe combinations are designed to detect. Thus, it is important to understand the scope of the assay’s detection when interpreting results. Also, different wine matrices can have a profound effect on DNA recovery. It is critical to know that your methods for DNA extraction and PCR are validated across all wine styles, and that proper controls are in place and used to detect any problems with DNA extraction and/or the PCR assay.

In fact, a significant improvement in PCR applications has been achieved with the Scorpions system from ETS. While it may not be practical for the small producer, currently there are three large wineries using the Scorpions technology in-house—two of them for more than two years—and several others that are planning to use the system soon. For high-volume producers, running PCR in-house is actually cheaper than sending samples out for plating.

One other interesting method that has been suggested for Brett detection is LAMP (Loop-mediated isothermal AMPlification). This is an affordable method that amplifies DNA with high specificity, efficiency and speed.4 While it has been shown to be useful,5 there apparently has not been much published on its application to wine, so its future use is unclear.

Plating and microscopy—new tricks
The group of yeasts currently called Brettanomyces bruxellensis is actually a number of similar yeasts that we simply call “Brett,” with a very wide range of characteristics.6 Since traditional agar plating usually is based on limiting growth to a fairly narrow range of characteristics, this creates some problems when trying to isolate and identify such a large group.

For example, some commercial media have been based on Brett’s ability to use ethanol as its sole carbon source, but a high percentage of strains—perhaps more than half—can’t do this. Brettanomyces bruxellensis is known to be highly resistant to cycloheximide (sold commercially as Actidione), which can be a useful factor, but quite a few other non-Saccharomyces wild yeasts are also resistant. In addition, it should be present at an adequate concentration, but WL-D, a popular formula, contains only 4ppm cycloheximide, probably not enough to inhibit many otherwise sensitive yeasts.

The addition of p-coumaric acid to the medium7a,b,c is a clever trick to provide a precursor to 4-ethylphenol, which is the major offensive odor of Brett spoilage, but not all Brett strains produce 4-EP (a few bacteria have been found to produce 4-EP, but only in juice and at fairly low concentrations8). Adding an acid indicator (see Figure 1 at top) can be helpful because most Bretts are vigorous acid producers, but this varies among strains, and other wild yeasts may also produce a lot of acid.

However, using a combination of these factors can produce a medium that is more likely to be useful for isolating Brett than relying on one or two factors alone. While not all Bretts will generate 4-EP from the p-coumaric acid additive, those that don’t may not be as great a concern for spoilage; other factors such as additions of antibiotics, lower pH and mold inhibitors help selective growth of Brett. The fact that Brett is one of the few yeast types that use cellobiose (abundant in new barrels) as a carbon source should also be of use in developing a better agar medium.

Table 1 compares several agar media for the isolation of Brett in wine.

Limitations of plating
Some amount of plating is a procedure that is usually well within the capabilities of most small wineries and should be for all larger ones. Costs are low, as pre-poured plates can be purchased, ovens or boiling should be adequate for sterilization, incubators are usually not necessary, and any winery should be embarrassed to be without a microscope.

However, considering 1) the speed at which Brett can take off in an individual tank or barrel and 2) the presence of dozens of tanks or thousands of barrels in a winery, plating becomes a Sisyphean task. Sampling alone could take a prohibitive amount of time, considering that the results are only good until the winemaker’s next sampling—or that winery’s next contamination.

The most practical solution to this is to sample at the time of blending or tank-filling. However, this inevitably dilutes the sample possibly past the point of detection, simultaneously increases the chances of interference from other organisms and still tests only short windows of time during the vinification process.

Moreover, if it takes a week or more to get visible growth of colonies, the wine could easily have become spoiled by the time the winemaker finds out about the contamination.

Another significant problem with plating, particularly with Brett, is the phenomenon of Viable But Not Culturable (VBNC) cells, also called NICs or Not Immediately Culturable. These are essentially dormant yeasts that don’t readily grow on plates under normal conditions. They are also unusually small cells, small enough to actually pass through a 0.45µm filter in some cases. This is a disturbing discovery, and it may be the reason for both some processing and plating failures. This is where PCR-based analysis would be clearly superior to plating.

Some new approaches
There are now two quite different alternate approaches to overcoming these limitations. The first is based on a theoretical application of current microbiological methods, and the second is a proven method that involves going straight to the source of the problem.

Epifluorescence micrograph Brettanomyces bruxellensis
Figure 2: Epifluorescence micrograph of Brettanomyces bruxellensis found on grapes. Arrows indicate viable cells. Renouf & Lonvaud-Funel, 2007
Epifluorescence microscopy is now being used to obtain immediate microscopic evaluation of wine samples9a,b,c by using fluorescent dyes to label both viable and non-viable (dead) cells (see Figure 2, at left). Obviously, the user can also easily distinguish between yeasts, bacteria and other materials in the sample at the same time, and cell counts can be made for evaluating inocula and fermentation status.

Another technique now widely used in medical diagnostics and food testing is the application of immunological methods. These are based on the use of antibodies to specific microorganisms, and provide a “magic bullet” to find small numbers of target organisms in the sample. By using an antibody specific to Brett yeasts bound to a fluorescent dye, one might be able to almost instantly identify microscopically any Brett cells present in a sample—cells that might otherwise be essentially invisible.

Previously the problem with obtaining antibodies to wine microorganisms was cross-reactions with other bugs, but recently Dr. Stewart Lebrun of Lebrun Labs in Irvine, Calif., developed a procedure to obtain anti-Brett antibodies from the yeast cell surface that have been shown in repeated testing to be both specific to Brett (with no significant cross reactions with other wine yeasts or bacteria) and versatile in recognizing a large number of Brett strains from both the United States and Europe. While not currently commercially available, I’m hopeful that there will be future interest in such an approach to detection of Brett and many other wine microorganisms such as Zygosaccharomyces bailii.

An ounce of prevention
A completely different approach to preventing Brett problems has been suggested by several research groups in France. This deals with finding the source of the problem before it even hits the crusher—that is, in the vineyard.

Many winemakers will testify to the tendency of certain vineyards or lots to be associated with Brett problems, which offers a preventative measure: If problem sources can be identified before they reach the fermentors, the contamination can be isolated and eliminated by use of centrifugation and/or extra SO2 before other lots are affected. Two research reports in recent years have devised some clever application of basic microbiology to identify problem spots in the vineyard that are the sources of Brett contamination.9a, 10

Renouf and Lafon-Lafourcade used a selective enrichment broth (EBB) inoculated with berries from suspected sites.9a This medium strongly favors Brett and is restrictive to most other yeasts, so that at 10-12 days, if Brett is present, it will be the overwhelmingly predominant yeast (108-109 cells/mL vs. 105-106 for other yeasts) and easily detected microscopically or by plating (this is done well before harvest, so the lag time for colony growth is not a problem). This could also be done using PCR/Scorpions methods as well, skipping the plating altogether.

Several years ago, while at Constellation, I was able to confirm by using this method one winemaker’s suspicion of a vineyard lot that had repeatedly been the apparent source of Brett at his winery. Results were confirmed with plating and immunological tests. Nearby lots that had not been a Brett problem gave negative results for Brett. It seems that this could be a very practical approach to dealing with the source of the Brett problem and go a long way towards successfully reducing the incidence of phenolic spoilage. It can also be done prior to harvest, at a time when the workload is a little lighter.

Brett spoilage is a preventable problem. Less adherence to stylistic dogmas would go a long way towards resolving this issue, but as long as winemaking methods coincide with ideal conditions for production of Brett, problems will persist unless some compromises are made. Until that happens, early detection is the best hope for minimizing its effect.

William Edinger, Ph.D., has worked in the wine industry for more than 30 years in New York, Ontario and California.  He was senior research microbiologist at Constellation Wines U.S. for 12 years and currently lectures about enology at at California State University, Fresno. To comment on this article, e-mail jim@winesandvines.com.

1. Renouf, V. & Lonvaud-Funel, A. 2007. Microbiol Res 162:154-167.
2. Stender, H., Kurtzman, C., Hyldig-Nielsen, J., Sørensen, D., Broomer, A., Oliveira, K., Perry-O’Keefe, H., Sage, A., Young, B., & Coull, J. 2001. Appl Envir Microbiol 67:938-941.
3. Phister, T. & Mills, D. 2003. Appl Envir Microbiol 69:7430-7434.
4. Notomi, T., Okayama, H., Masubuchi, H., Yonekawa, T., Watanabe, K., Amino, N., & Hase, T. 2000. Nucleic Acids Res 28(12):e63.
5. Hayashi, N., Arai, R., Tada, S., Taguchi, H., & Ogawa, Y. 2007. Food Microbiol 24:778-785.
6. Conterno, L., Joseph, L., Arvik, T., Henick-Kling, T., & Bisson, L. 2006. Am J Enol Vitic 57:139-147.
7a. Benito, S., Palomero, F., Morata, A., Calderón, F., & Suàrez-Lepe, J. 2009. J Appl Microbiol 106:1743-1751.
7b. Rodrigues, N., Gonçalves, G., Pereira-da-Silva, S., Malfeito-Ferreira, M., & Loureiro, V. 2001. J Appl Microbiol 90:588-599.
7c. Couto, J., Barbosa, A., & Hogg, T. 2005. L Appl Microbiol 41:505-510.
8. Couto, J., Campos, F., Figueiredo, A., & Hogg, T. 2006. Am J Enol Vitic 57:166-171.
9a. Renouf, V., Lonvaud-Funel, A., & Coulon, J. 2007. Intl J Vine Wine Sci 41:161-173.
9b. Martens-Habbena, W. & Sass, H. 2006. Appl Environ Microbiol 72:87-95.
9c. Kopke, C. Cristovão, A. Prata, A.,Silva Pereira, C., Figueiredo Marques, J., & San Rom‹o, M. 2000. Food Microbiol 17:257-260.
10. Barbin, P., Strehaiano, P., Taillandier, P., & Gilis, J. 2007. Rev des Oenologues No. 124, July 2007 [in French].
Print this page   PRINTER-FRIENDLY VERSION   »
E-mail this article   E-MAIL THIS ARTICLE   »
Currently no comments posted for this article.