Detecting an aroma defect before it even becomes noticeable in wine, assessing the water stress of a plot, or tracing the origin of its nitrogen nutrition: behind their image of cutting-edge technology, isotopes find very practical applications in the wine industry. From mass spectrometry to isotope dilution, they are becoming genuine diagnostic tools to support a viticulture industry facing new challenges.

Climate change, evolving raw materials, reduction of inputs, adaptation of oenological practices, demand for lower-alcohol wines... The wine sector must now cope with multiple, sometimes conflicting constraints. For laboratories, these transformations raise an essential question: how can we better understand what is happening in the vineyard and in the wine in order to intervene at the right time?

During a session presented at Forum LABO, researchers from the UMR Sciences for Oenology and the EXCELL laboratory showed how isotopic analysis could provide part of the answer.

With more than a thousand identified aroma compounds, of which about 10% are considered to have an impact on aroma, wine is a particularly complex matrix to analyze. Some of these compounds are present in trace amounts and can be extremely reactive. To detect them, analytical methods must therefore achieve very high levels of sensitivity. This challenge becomes particularly critical when it comes to identifying an aromatic fault.

Among them is the fresh mushroom aroma (FMA), associated with fungal spoilage of the raw material and recurrently observed in various wine-growing regions, making its detection particularly delicate.

The defect may be visible on the cluster, become barely noticeable in the must, and then reappear at the end of alcoholic fermentation. Its intensity can then continue to fluctuate in the wine. In other words, waiting until you can smell the defect in the finished product may already be too late.

Several families of molecules were already associated with this defect, particularly certain C7, C8, and C9 compounds. Among them, oct-1-en-3-one is considered particularly impactful on fresh mushroom notes. However, its use as a marker has a major limitation: its detection occurs mainly in the final product and relies on relatively time-consuming GC-MS analysis. The research presented during the session led to the identification of two new markers: hydroxyoctanone and octane-1,3-diol.

The second was notably already known in apples, but was identified for the first time in wines presenting the defect under study. The challenge then became analytical: how to quantify these compounds in a sensitive, reliable, rapid, and sufficiently robust manner to consider transferring to industrial partners?

To address this issue, researchers turned to a gold-standard technique in mass spectrometry: isotope dilution. The principle involves adding a known quantity of an internal standard to the sample that has an almost identical structure to the target molecule, but is labeled with stable isotopes. In the case presented, certain hydrogen atoms are, for example, replaced by deuterium. The major advantage is that the analyte and its isotopically labeled analogue behave almost identically during the analysis.

If part of the target molecule is lost or degraded during sample preparation, its internal standard will theoretically undergo the same phenomenon. Quantification is then based on the ratio between the two signals, which corrects for a large portion of analytical biases and matrix effects. However, this robustness comes with a trade-off: isotopically labeled standards are not always commercially available and may require real expertise in organic synthesis. The positioning of the isotopes must also be sufficiently stable to prevent exchange or loss during analysis.

The developed method combines liquid chromatography with a triple quadrupole mass spectrometer, used in MRM mode. This approach makes it possible to precisely select the ion corresponding to the target compound, fragment it, and then identify specific characteristic fragments. Multiple transitions are used to confirm the identity of the molecule and enhance measurement selectivity. However, one of the key challenges was also to minimize sample preparation as much as possible.

For wine, the presented protocol is particularly simple: 1 mL of wine, addition of internal standards, centrifugation, then injection. The chromatographic separation is performed on a C18 reversed phase with a 16-minute gradient. The method was then evaluated according to several criteria: linearity, matrix effect, accuracy, repeatability, reproducibility, and limits of detection and quantification.

The presented results notably show accuracies between 90 and 110% and repeatability and reproducibility levels below 15%. Furthermore, isotope dilution allows for a very strong correction of the matrix effects initially observed.

This is probably where the application becomes most valuable for the industry. Around sixty wines were analyzed, about half of which showed no ACF defect during sensory evaluation, while the other half exhibited varying levels of contamination. The concentrations of the new markers made it possible to significantly distinguish between the two groups and showed a strong correlation with the sensory intensity of the defect.

Most importantly, the researchers reproduced this discrimination directly in the musts. Twenty musts were analyzed: ten that subsequently yielded a defect-free wine and ten that led to a wine exhibiting an ACF. Here again, a significant difference was observed. These results pave the way for an early diagnosis, even before the defect clearly appears in the final product. Ultimately, the analysis could complement the sanitary assessment of grapes and make it possible to classify tanks according to their risk level, in order to adapt winery decisions earlier.

Another advantage of this approach: it is not limited to the oenological matrix. The protocol has also been adapted to apple juice, simply with an additional grinding and filtration step. In this matrix, the same compounds are no longer associated with a fresh mushroom aroma, but with what can be described as a cellar taste. Here again, the method made it possible to distinguish control samples from contaminated samples. An illustration of the value of multi-matrix analytical methods, capable of being deployed across several agri-food sectors.

The mechanism relies directly on photosynthesis. When the grapevine has sufficient water, the stomata on the leaves remain open, allowing gas exchange. CO₂ enters the leaf, and Rubisco, a key enzyme in photosynthesis, naturally favors the lighter carbon-12 over carbon-13. Under water stress conditions, the plant progressively closes its stomata to limit water loss.

The amount of CO₂ available inside the leaf then decreases. The discrimination between carbon-12 and carbon-13 becomes less significant, and a higher proportion of carbon-13 ends up in the produced sugars. By subsequently analyzing the isotopic ratio of these sugars in the grape, it becomes possible to determine the average level of water stress experienced by the vine.

The advantage of δ¹³C is that it constitutes an integrative indicator. A single measurement taken close to harvest provides information that integrates the conditions experienced by the plant during the sugar accumulation period. It is therefore not a tool designed for real-time irrigation management, but rather a way to understand the overall behavior of the vine over a given period.

The data presented by the EXCELL laboratory is based on a database of more than 10,000 samples. On a national scale, they notably highlight a vintage effect. A hot and dry vintage thus presents a different profile from a much rainier year. But the narrower the scale of observation—region, appellation, estate, then parcel—the more precise the information becomes.

Isotope analysis then takes on a particularly practical dimension. By combining the results with plot coordinates, it becomes possible to build a mapping of δ¹³C at the scale of a wine estate. Winegrowers can thus visualize the areas that have experienced the most significant water stress and compare their behavior from one vintage to another. These differences may notably be linked to soil types.

Clay soils, capable of retaining more water, can for instance display different profiles compared to sandier or gravelly soils. The presence of limestone can also play a role by promoting the storage and then gradual release of water. Isotope analysis then becomes an additional tool to understand why two plots of the same estate do not react in the same way to the same climatic conditions.

Carbon is not the only element capable of providing information. The nitrogen isotope ratio, or δ¹⁵N, makes it possible to explore another dimension: that of nitrogen nutrition and the origin of the organic matter used by the plant. Fertilizers and soil improvers can indeed display different isotopic signatures depending on their origin. The data presented thus show differences between vineyards primarily using mineral fertilization and others resorting to organic amendments. This signature can then be traced all the way into the plant and the grape. Still emerging in viticulture, δ¹⁵N analysis could therefore complement the information provided by δ¹³C and contribute to a more detailed understanding of how plots function.

At first glance, the two approaches presented seem very far apart. On the one hand, isotopically labeled standards enable the precise quantification of trace molecules and the anticipation of aromatic defects. On the other hand, natural variations in carbon and nitrogen isotopes become markers for the functioning of the vine and its environment.

Yet, they are based on the same idea: using isotopes as tracers capable of revealing information invisible to the naked eye.

In an industry facing climate change, evolving consumer expectations, and the need to better manage resources, this ability to detect, understand, and anticipate could become particularly valuable.

From the grape to the plot, and from the molecule to the vintage, isotopes thus allow the laboratory to read part of the history of the vine and wine.

Article based on the session dedicated to isotope applications in the wine sector, presented during Forum LABO.