Sequential Non-Saccharomyces and Saccharomyces cerevisiae Fermentations to Reduce the Alcohol Content in Wine

Round 1

Reviewer 1 Report

The work “Sequential non-Saccharomyces and Saccharomyces cerevisiae Fermentations to Reduce the Alcohol Content in Wine” is of  interest is of actual and applicative interest in the sector of wine production, more and more addressed to characterize the and differentiate the wines, also with the aim to implement the use of selected starter in order to achieve specific aims. This work is focused on the selection of non-Saccharomyces yeasts to use in sequential fermentation with Saccharomyces cerevisiae in order to reduce the ethanol content of the final products. This topic is of wide discussion for both researchers and wine producers and in this work the authors present new results which, after confirmation on a pilot scale, seem promising for use in the cellar.

The experimental design is correct and extensive and the methods are clearly described in detail. The authors report results, which are  presented in tables and figures and  statistically elaborated,  thus producing an immediate understanding to the reader regarding the correlation between the experimental wine parameters in function of the different non-Saccharomyces in the sequential fermentations. I suggest the authors to add a figure for the results described in the lines 156-170.

This paper merits to be published after minor revision.

Comments:

Line 139 …….. which have shown similar fermentative behavior to the control at high level of sugars in aerobic conditions??  It is not clear. Perhaps the authors intend: …..which have shown similar CO₂ released to Sc CLI 889 control strain?

Table 3 Ethyl acetate has not been determined ??? It is the ester present in the greatest quantity in wine

Fig 2 PCA-A Why in the PCA are some combinations listed in duplicate and others not? In particular:

Wa 21A-5C/Sc CLI 889  and Sc CLI 889 once,whereas Mp CLI 68/Sc CLI 889, Mg CLI 1217/Sc CLI 889 and Mp CLI 460/Sc CLI 889  twice.

Fig 2 PCA B   It is clearer to quote in fig 2: ∑Alcohols, ∑Esters, ∑Acids, ∑Aldehydes/Ketones, as reported in table 3

Author Response

The changes suggested by reviewers have been highlighted in yellow in the revised manuscript.

Comments and Suggestions for Authors

The work “Sequential non-Saccharomyces and Saccharomyces cerevisiae Fermentations to Reduce the Alcohol Content in Wine” is of  interest is of actual and applicative interest in the sector of wine production, more and more addressed to characterize the and differentiate the wines, also with the aim to implement the use of selected starter in order to achieve specific aims. This work is focused on the selection of non-Saccharomyces yeasts to use in sequential fermentation with Saccharomyces cerevisiae in order to reduce the ethanol content of the final products. This topic is of wide discussion for both researchers and wine producers and in this work the authors present new results which, after confirmation on a pilot scale, seem promising for use in the cellar.

The experimental design is correct and extensive and the methods are clearly described in detail. The authors report results, which are presented in tables and figures and statistically elaborated, thus producing an immediate understanding to the reader regarding the correlation between the experimental wine parameters in function of the different non-Saccharomyces in the sequential fermentations. I suggest the authors to add a figure for the results described in the lines 156-170. We did not include it on manuscript because we did not consider necessary it to understand the work. Figure 1 help us to explain the clear classification of yeast fermentative profiles in three different groups from the control during pure culture. After sequential fermentation (page 6, lines 153‒161), we based our yeast selection focusing on oenological parameters (residual sugar or alcohol content), and not so CO₂ released. However, if you consider that this information should be added in the manuscript, feel free to propose to include it.

This paper merits to be published after minor revision.

Comments:

Line 139 …….. which have shown similar fermentative behavior to the control at high level of sugars in aerobic conditions??  It is not clear. Perhaps the authors intend: …..which have shown similar CO₂ released to Sc CLI 889 control strain?

We have clarified the sentence to ensure that no possible confusion could be created. Page 4, lines 139‒140 …we added the phrase: “which have shown similar CO₂ released to S. cerevisiae CLI 889 control strain”

Table 3 Ethyl acetate has not been determined??? It is the ester present in the greatest quantity in wine

Ethyl acetate concentration was below its odour threshold in all wines using our analysis method, so this compound has not been included on Table 3.  

Fig 2 PCA-A Why in the PCA are some combinations listed in duplicate and others not? In particular:

Wa 21A-5C/Sc CLI 889 and Sc CLI 889 once, whereas Mp CLI 68/Sc CLI 889, Mg CLI 1217/Sc CLI 889 and Mp CLI 460/Sc CLI 889 twice.

All combinations are in duplicate. In case of Wa 21A-5C/Sc CLI 889 and Sc CLI 889 control, oenological and aroma values classified the two samples very close and they are overlapped.

Fig 2 PCA B   It is clearer to quote in fig 2: ∑Alcohols, ∑Esters, ∑Acids, ∑Aldehydes/Ketones, as reported in table 3

Figure 2B have been changed. The symbol ∑ has been added to represent total alcohols, esters, acids and aldehydes/ketones.

Author Response File: Author Response.docx

Reviewer 2 Report

In this study, the authors aim to address a challenge in the wine industry, that is producing wines with lower ethanol content. The authors choose a microbiological approach that has recently gained momentum e.g. co-cultures of S. cerevisiae and other yeast species. Here, the authors perform a sequential culture approach, starting with a fermentation using native non-Saccharomyces yeasts followed by addition of S. cerevisiae.

During the first 96h in a pure non-Saccharomyces culture, the strains resulted, as can be expected from the diversity of chosen species, in vastly different amounts of ethanol content from low to equivalent amounts to that of the S. cerevisiae control. The authors then inoculated S. cerevisiae in these 96h fermentations and observed that many combinations resulted either in a final ethanol concentration similar to that of the control (and therefore unsuited for the purpose of lower-ethanol winemaking) or to high amounts of residual sugars. However, four combinations worked better, with a final alcohol content of 11.75%-12.16% (vs 13% control). A volatile compound analysis is then presented; there is a significant difference in the profile of the sequential fermentations vs the controls, indicating that the resulting wine may display different qualities to that of the only-S. cerevisiae control—these qualities are however hard to assess subjectively based simply on a volatile profile. Overall, this study presents a proof-of-concept that a sequential fermentation approach is possible using native yeasts of this region. As noted by the authors, a larger-scale test as well an evaluation of the sensorial profile of the obtained wines will be important to know whether this process can present an answer to the challenge of ethanol content in winemaking and be a technologically and/or commercially viable development.

The data is fairly straight-forward as presented in the paper. In section 3-2, line 165, it would be interesting to specify which rejected strains belong to group (B) and which to group (C).

Moreover, in Figure 1, especially Fig 1C, some strain markers are indistinguishable from each other (virtually identical dot types, same color). This raises the question of whether displaying the data in this fashion is the clearest way to do so. I would suggest either using colors, or splitting the group C into several graphs with better readability. An alternative could be to display on a single graph the control, the average of group (A), the average of group (B), and the average of group (C) (maybe also with a shaded region indicating the in-group variance), with different colors, and then refer to a supplemental figure or table for the precise values for each strain.

Author Response

The changes suggested by reviewers have been highlighted in yellow in the revised manuscript.

Comments and Suggestions for Authors

In this study, the authors aim to address a challenge in the wine industry, that is producing wines with lower ethanol content. The authors choose a microbiological approach that has recently gained momentum e.g. co-cultures of S. cerevisiae and other yeast species. Here, the authors perform a sequential culture approach, starting with a fermentation using native non-Saccharomyces yeasts followed by addition of S. cerevisiae.

 During the first 96h in a pure non-Saccharomyces culture, the strains resulted, as can be expected from the diversity of chosen species, in vastly different amounts of ethanol content from low to equivalent amounts to that of the S. cerevisiae control. The authors then inoculated S. cerevisiae in these 96h fermentations and observed that many combinations resulted either in a final ethanol concentration similar to that of the control (and therefore unsuited for the purpose of lower-ethanol winemaking) or to high amounts of residual sugars. However, four combinations worked better, with a final alcohol content of 11.75%-12.16% (vs 13% control). A volatile compound analysis is then presented; there is a significant difference in the profile of the sequential fermentations vs the controls, indicating that the resulting wine may display different qualities to that of the only-S. cerevisiae control—these qualities are however hard to assess subjectively based simply on a volatile profile. Overall, this study presents a proof-of-concept that a sequential fermentation approach is possible using native yeasts of this region. As noted by the authors, a larger-scale test as well an evaluation of the sensorial profile of the obtained wines will be important to know whether this process can present an answer to the challenge of ethanol content in winemaking and be a technologically and/or commercially viable development.

The data is fairly straight-forward as presented in the paper. In section 3-2, line 165, it would be interesting to specify which rejected strains belong to group (B) and which to group (C).

Page 6, lines 160‒171: we have specified if non-Saccharomyces strains named in this paragraph were previously classified in group B or C in the Section I (pure culture).

Moreover, in Figure 1, especially Fig 1C, some strain markers are indistinguishable from each other (virtually identical dot types, same color). This raises the question of whether displaying the data in this fashion is the clearest way to do so. I would suggest either using colors, or splitting the group C into several graphs with better readability. An alternative could be to display on a single graph the control, the average of group (A), the average of group (B), and the average of group (C) (maybe also with a shaded region indicating the in-group variance), with different colors, and then refer to a supplemental figure or table for the precise values for each strain.

The sense of Figure 1 is to show a trend in which the yeast strains are classified. As we described at the foot of the figure, we divided the studied strains in three groups in relation with their fermentative capacity (measured as weight loss) from the same control. In this case, we aim to generate a fast-visual classification, separating the strains with similar CO₂ released than control (A), intermediate CO₂ released (B) and low CO₂ released below 1 g/L and more information about the yeast name within each group can be found on the legend chart.

We have considered the reviewer’ proposals but we have found some problems:

• We cannot do the average of each group because each group contains different yeast species with fermentative variability inter-species and intra-species, particularly noticeable in group B.
• If splitting the group C into several groups, the CO₂ loss profiles remain very close. We showed below an attempt to divide group C into three graphs using colors.

  The changes suggested by reviewers have been highlighted in yellow in the revised manuscript.

   

  Comments and Suggestions for Authors

  In this study, the authors aim to address a challenge in the wine industry, that is producing wines with lower ethanol content. The authors choose a microbiological approach that has recently gained momentum e.g. co-cultures of S. cerevisiae and other yeast species. Here, the authors perform a sequential culture approach, starting with a fermentation using native non-Saccharomyces yeasts followed by addition of S. cerevisiae.

   During the first 96h in a pure non-Saccharomyces culture, the strains resulted, as can be expected from the diversity of chosen species, in vastly different amounts of ethanol content from low to equivalent amounts to that of the S. cerevisiae control. The authors then inoculated S. cerevisiae in these 96h fermentations and observed that many combinations resulted either in a final ethanol concentration similar to that of the control (and therefore unsuited for the purpose of lower-ethanol winemaking) or to high amounts of residual sugars. However, four combinations worked better, with a final alcohol content of 11.75%-12.16% (vs 13% control). A volatile compound analysis is then presented; there is a significant difference in the profile of the sequential fermentations vs the controls, indicating that the resulting wine may display different qualities to that of the only-S. cerevisiae control—these qualities are however hard to assess subjectively based simply on a volatile profile. Overall, this study presents a proof-of-concept that a sequential fermentation approach is possible using native yeasts of this region. As noted by the authors, a larger-scale test as well an evaluation of the sensorial profile of the obtained wines will be important to know whether this process can present an answer to the challenge of ethanol content in winemaking and be a technologically and/or commercially viable development.

  The data is fairly straight-forward as presented in the paper. In section 3-2, line 165, it would be interesting to specify which rejected strains belong to group (B) and which to group (C).

  Page 6, lines 160‒171: we have specified if non-Saccharomyces strains named in this paragraph were previously classified in group B or C in the Section I (pure culture).

  Moreover, in Figure 1, especially Fig 1C, some strain markers are indistinguishable from each other (virtually identical dot types, same color). This raises the question of whether displaying the data in this fashion is the clearest way to do so. I would suggest either using colors, or splitting the group C into several graphs with better readability. An alternative could be to display on a single graph the control, the average of group (A), the average of group (B), and the average of group (C) (maybe also with a shaded region indicating the in-group variance), with different colors, and then refer to a supplemental figure or table for the precise values for each strain.

  The sense of Figure 1 is to show a trend in which the yeast strains are classified. As we described at the foot of the figure, we divided the studied strains in three groups in relation with their fermentative capacity (measured as weight loss) from the same control. In this case, we aim to generate a fast-visual classification, separating the strains with similar CO₂ released than control (A), intermediate CO₂ released (B) and low CO₂ released below 1 g/L and more information about the yeast name within each group can be found on the legend chart.

  We have considered the reviewer’ proposals but we have found some problems:

  • We cannot do the average of each group because each group contains different yeast species with fermentative variability inter-species and intra-species, particularly noticeable in group B.
  • If splitting the group C into several groups, the CO₂ loss profiles remain very close. We showed below an attempt to divide group C into three graphs using colors.

Author Response File: Author Response.docx

Reviewer 3 Report

Dear editor and authors,

In the present work, the authors have tried to develop new strategies to reduce the ethanol content in wines made by the variety of Malvar, by co-inoculating non-Saccharomyces yeast strains with one Saccharomyces cerevisiae strain. Besides the ethanol content in the final wines, the authors have analyzed important oenological parameters and volatiles compounds. PCA analysis with the tested parameters and metabolites was performed in the end. The text is well written and documented. My main concern is the selection of one control strain of S. cerevisiae. I think is important to explain why they have chosen this strain for their experiment based on the fact that S. cerevisiae strains can exhibit great differences concerning their fermentation performances and their sensory contribution in wine. Additionally it could have been nice to finalize their work by a sensory evaluation of the produced wines especially in the cases where volatile acidity is relatively high.

You can read below more suggestions.

Line 13

The consumer preference is not linked with the sugar concentration in harvest

Line 25

I am not sure you can write about organoleptic characteristics as you didn’t realize any sensory evaluation of the final wines.

Introduction

You work focus a lot on non-Saccharomyces yeast. I think you should add more bibliography about their role in oenology in the introduction

Line 57

Can you explain more that phrase ‘’ new styles of wine more competitive in the market’’?

Line 58

What do you mean by knowledge?

Line 86

pH? Any sulphites added?

Line 87

How you estimate the final concentration? all the cultures after 48h had the same population?

Line 88

By which criterion you have chosen your control strain? Do you think is enough to use just one strain of  S. cerevisiae?

Line 92

Why you have chosen the moment of 96 hours for the inoculation of S. cerevisiae?

Line 96

The fermentation process in your research was under shaking but in real winemaking conditions that doesn’t happen. How you can explain your decision?

Table 2

I am not sure that dry weight can be considered as an oenological parameter

Figures

Pay attention to the quality of the figures!!

Line 272

How according to your opinion the production of acetic acid can be reduced?

Author Response

The changes suggested by reviewers have been highlighted in yellow in the revised manuscript.

Comments and Suggestions for Authors

Dear editor and authors,

In the present work, the authors have tried to develop new strategies to reduce the ethanol content in wines made by the variety of Malvar, by co-inoculating non-Saccharomyces yeast strains with one Saccharomyces cerevisiae strain. Besides the ethanol content in the final wines, the authors have analyzed important oenological parameters and volatiles compounds. PCA analysis with the tested parameters and metabolites was performed in the end. The text is well written and documented. My main concern is the selection of one control strain of S. cerevisiae. I think is important to explain why they have chosen this strain for their experiment based on the fact that S. cerevisiae strains can exhibit great differences concerning their fermentation performances and their sensory contribution in wine. Additionally, it could have been nice to finalize their work by a sensory evaluation of the produced wines especially in the cases where volatile acidity is relatively high. We are totally in agreement with the reviewer, future work will be directed to test these preliminary results at pilot scale in cellar. The greater volume of wine will permit us to confirm yeast strains combinations as low-ethanol producers and we could describe sensorially these wines.

You can read below more suggestions.

Line 13. The consumer preference is not linked with the sugar concentration in harvest.

These consumer preferences are related to rich, full-bodied and ripe fruit flavor profiles and less for unripe green and vegetal wine flavors. The production of the riper styles of wine is achieved by extending the time before harvest and therefore increasing grape sugar content.

However, we have nuanced the sentence, including: “consumer preferences for particular wine styles” (page 1, line 13).

Line 25. I am not sure you can write about organoleptic characteristics as you didn’t realize any sensory evaluation of the final wines.

Page 1, line 25: we have changed the word “organoleptic” by “oenological”. In this sense, we confirm that sequential combinations impart good oenological characteristics to wines as high glycerol proportion, volatile higher alcohols and esters with fruity and sweet character but not tested by sensorial panel.

Introduction. You work focus a lot on non-Saccharomyces yeast. I think you should add more bibliography about their role in oenology in the introduction.

We consider that the relation between non-Saccharomyces and the reduction of alcohol content in wine is well documented in the manuscript. The use of non-Saccharomyces is being broadly studied in winemaking nowadays but we focus the introduction information with the main topic. However, if you consider that this information should be important to manuscript understanding, feel free to propose to include it.

Line 57. Can you explain more that phrase ‘’ new styles of wine more competitive in the market’’?

The D.O. “Vinos de Madrid” main objectives are to stimulate the production of high quality young and aged wines and to foster the foreign exports. All these trends bring up the need to get deeper in the characterization of D.O. “Vinos de Madrid” wines, increasing the number of representative samples and extending the periods of follow-up for their better classification and differentiation. In this sense, winemakers are taking an intense labor of modernization and investigation, including tasks aimed to increase the knowledge about the native yeast communities present in vineyards and wineries as well as their application to elaborate products with their own typicity.

Line 58. What do you mean by knowledge?

We have included the wide word “knowledge” to express the importance of a deep understanding about the fermentative requirements and metabolism of yeasts as well as their chemical and sensory impact in wine.

Line 86. pH? Any sulphites added?

The pH value of Malvar must was 3.3, this value is included in the manuscript (page 2, line 84).

No additional sulphites source was added in the must. As we commented in the discussion (page 10, lines 290‒292), maybe nutritional requirements by S. cerevisiae caused the elevated volatile acidity values at the end of fermentation. We should consider this point, adding a nitrogen source in future work.

Line 87. How you estimate the final concentration? all the cultures after 48h had the same population?

The cell concentration in pre-cultures were calculated after optical density determination at 600 nm.

After 48h, pre-cultures of yeast strains had different cell concentration but we calculated the volume of pre-culture needed to start each fermentation in 50 mL with 10⁶ cells/mL.

Line 88. By which criterion you have chosen your control strain? Do you think is enough to use just one strain of S. cerevisiae?

As referenced in the manuscript, this native S. cerevisiae strain from D.O. “Vinos de Madrid” has been isolated and widely studied by our research group. We selected S. cerevisiae CLI 889 as control due to its good oenological aptitudes and its improvement of Malvar wines quality. Moreover, we promote the use of this autochthonous Saccharomyces strain in winemaking, which can impart personality to Malvar wines in this production area.

Line 92. Why you have chosen the moment of 96 hours for the inoculation of S. cerevisiae?

At this time, the first part of fermentations (Section I, pure culture) had finished, since the weight loss was stabilized.

Line 96. The fermentation process in your research was under shaking but in real winemaking conditions that doesn’t happen. How you can explain your decision?

Due to the small volumes used, agitation is important to mix well all the content and provide the dissolved oxygen required for the development of yeast biomass. Industrial fermentations do not need this agitation because the CO₂ produced during fermentation helps the mixing process of the must, however with these small volumes has been showed a good practice to reproduce the industrial situations and prevent the yeast settling.

Table 2. I am not sure that dry weight can be considered as an oenological parameter.

Page 6, line 172: we are changed the Table 2 caption by “Table 2. Oenological parameters and cell dry weight for the best non-Saccharomyces/S. cerevisiae sequential combinations to reduce ethanol concentration in wines”

Figures. Pay attention to the quality of the figures!!

We have checked the figures quality. These figures had been included in the article as vectorial images to ensure their correct view.

Line 272. How according to your opinion, the production of acetic acid can be reduced?

As we commented in page 10, lines 287−289, volatile acidity values were increased after S. cerevisiae inoculation in most cases. One cause could have been due to low levels of nutrients available to S. cerevisiae. This problem could be minimized by adding nutrients at same time to S. cerevisiae inoculation. In our experience with different fermentation volumes, wines fermented at high grape must volume help control low volatile acidity values. Future work will be directed towards the application of these practices.

Author Response File: Author Response.docx