Screening and Engineering Yeast Transporters to Improve Cellobiose Fermentation by Recombinant Saccharomyces cerevisiae

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have shown that expressing an intracellular beta-glucosidase BGL7 from Spathaspora passalidarum and the cellobiose transporter CBT2 from Meyerozyma guilliermondii in S. cerevisisae allow for the effecient fermentation of cellobiose, an important way to convert agricultural waste into bioful. They also show that not all heterologous cellobiose transporters are functional in cerevisiae and that CBT2 is recognized by the ubiquitination machinery that down-regulates plasma membrane proteins.

I'm missing in the paper the information about whether cellobiose is naturally occuring or the result of the enzymatic pre-treatment required for yeast to ferment cellulose. This is important to understand the results of figure 6 in which a mix of xylose and cellbiose, probably reflecting the situation industrials will be confronted with.

In analyzing all their transporter constructs, the author never include the growth rate but only the conversion of cellobiose into ethanol. But a good indicator of functionality of a heterologous transporter is the growth rate.

My biggest issue is with the interpretation of figure 2. The author conclude that the fermentation is stuck because 16g/L are converted into 8g/L of ethanol. I disagree because 16g of cellobiose represents 46.7 mM and 8g of ethanol represent 174 mM. Cellobiose if I'm not mistaken is the equivalent of 4 ethanol molecules hence 46.7 mM of cellobiose would represent 186 mM of ethanol if conversion were total. The full length CBT2would allow for 93% conversion whic is far from stuck in my opinion.

Because of their interpretation of the first conversion results, the authors made CBT2 truncation to relieve the perceived problem of internalization which would limit the conversion. To guide them they used 2 prediction websites. Unfortunately, one is no longer accessible and the other was down for debugging purposes. Based on these results they concluded that ubiquitination could occur at the N- and C-termini of CBT2 but only tested the C- or the N- and C-terminal deletions. The former giving similar results to the full length and the latter an improvement. Their conclusion is that both ubiquitination sites have to be removed to prevent internalization. This is in contradiction to what has been observed for many transporters in which only the N-terminus is ubiquitinated. The prediction site http://old.protein.bio.unipd.it/rubi/ suggests also that only the N-terminus is ubiquitinated. However it is true at least for Gap1 that the C-terminus contains signal(s) important its proper internalization. Furthermore the Y(p/s) for the full length is better than for the truncated forms which supports that the ubiquitination is not the reason for the stuck fermentation.

I think it would be wise to reperform the experiment in figure 6 with the full length transporter to see if this results in an improvement of conversion.

Finally, the authors discuss the phylogeny of beta-glucosidase without reference or tree for support. In figure 7 the authors propose a phylogenetic tree of cellobiose transporters and I would have expected to root it some well known hexose transporters instead of a putative Trichoderma reesei cellobiose transporter.

Author Response

Comments 1: The authors have shown that expressing an intracellular beta-glucosidase BGL7 from Spathaspora passalidarum and the cellobiose transporter CBT2 from Meyerozyma guilliermondii in S. cerevisisae allow for the effecient fermentation of cellobiose, an important way to convert agricultural waste into bioful. They also show that not all heterologous cellobiose transporters are functional in cerevisiae and that CBT2 is recognized by the ubiquitination machinery that down-regulates plasma membrane proteins.

Response 1: We would like to thank the reviewers for their comments/suggestions/criticisms, which we feel contributed to improve our manuscript. Detailed answers to the points raised by each reviewer are given below, and all modifications in the text are marked with yellow.

Comments 2: I'm missing in the paper the information about whether cellobiose is naturally occuring or the result of the enzymatic pre-treatment required for yeast to ferment cellulose. This is important to understand the results of figure 6 in which a mix of xylose and cellbiose, probably reflecting the situation industrials will be confronted with.

Response 2: We have added the following paragraph “Cellobiose can be found in small amounts in some fruits, honey or sugar beet, but the main source of this disaccharide is during cellulose degradation by microrganisms.” (L61-63 in the revised manuscript). And yes, a mix of xylose and cellobiose is produced during industrial 2G ethanol production, as discussed in L71-76 in the revised manuscript.

Comments 3: In analyzing all their transporter constructs, the author never include the growth rate but only the conversion of cellobiose into ethanol. But a good indicator of functionality of a heterologous transporter is the growth rate.

Response 3: We have included the growth rates for the strains expressing the 3 transporters (L349-352 in the revised manuscript), which are indeed very low! However, the reviewer should consider that all 3 cloned transporters are functional (they allow growth on cellobiose), but from an industrial point of view, the product (ethanol) is what is important, and only one permease allowed production of ethanol when expressed in S. cerevisiae.

Comments 4: My biggest issue is with the interpretation of figure 2. The author conclude that the fermentation is stuck because 16g/L are converted into 8g/L of ethanol. I disagree because 16g of cellobiose represents 46.7 mM and 8g of ethanol represent 174 mM. Cellobiose if I'm not mistaken is the equivalent of 4 ethanol molecules hence 46.7 mM of cellobiose would represent 186 mM of ethanol if conversion were total. The full length CBT2 would allow for 93% conversion which is far from stuck in my opinion.

Response 4: Since Reviewer # 2 also suggested removing “stuck” from the abstract, we have removed all “stuck fermentation” from the text, and used instead “incomplete fermentation” (see L24, L31, L104, L360, L366, L375, L386, L393, L398, L427, L442, L499, L501, L506, L543, and L579). “Stuck fermentation” is a term used to define fermentations were all the sugar is not completely consumed (very common in wine fermentations) and thus this sugar remains in the medium. We agree with the reviewer that the sugar that enters the cell is fermented efficiently (with a high yield of g ethanol per g of consumed sugar), but the problem is that at a certain point the cells stop consuming the sugar!

Comments 5: Because of their interpretation of the first conversion results, the authors made CBT2 truncation to relieve the perceived problem of internalization which would limit the conversion. To guide them they used 2 prediction websites. Unfortunately, one is no longer accessible and the other was down for debugging purposes. Based on these results they concluded that ubiquitination could occur at the N- and C-termini of CBT2 but only tested the C- or the N- and C-terminal deletions. The former giving similar results to the full length and the latter an improvement. Their conclusion is that both ubiquitination sites have to be removed to prevent internalization. This is in contradiction to what has been observed for many transporters in which only the N-terminus is ubiquitinated. The prediction site http://old.protein.bio.unipd.it/rubi/ suggests also that only the N-terminus is ubiquitinated. However it is true at least for Gap1 that the C-terminus contains signal(s) important its proper internalization. Furthermore the Y(p/s) for the full length is better than for the truncated forms which supports that the ubiquitination is not the reason for the stuck fermentation.

Response 5: Unfortunately the two prediction websites are indeed not working today (but when we use them there were OK). We did not understand the observation made by the reviewer that the Rubi website shows “that only the N-terminus is ubiquitinated”…. we used this website to analyze the CBT2 transporter, and the results with this website showed only one lysine (K267) had ubiquitination potential….!!! (a lysine in the intracellular loop between TM-6 and TM-7). This website did not recognized any of the N- or C-terminal lysine residues as possible ubiquitination sites. We first focused in the C-terminal domain because it had 3 lysine with high ubiquitination potential, while those at the N-terminal domain had medium ubiquitination potential. But as discussed in L504-521, there are several manuscripts showing that lysine residues at the C-terminal domain are involved in ubiquitination and dowregulation of sugar transporters expressed in S. cerevisiae (including the CDT-2 cellobiose transporter from N. crassa)… see ref. [69-72]. And again, one thing is the ethanol yield, another thing is incomplete sugar consumption from the medium.

Comments 6: I think it would be wise to reperform the experiment in figure 6 with the full length transporter to see if this results in an improvement of conversion.

Response 6: We sincerely do not think that the full length transporter will perform better during xylose-cellobiose co-fermentation (is already bad with cellobiose fermentation, not consuming all the sugar). Furthermore, the editor of Fermentation gave us just 10 days to re-submit our revised version of the manuscript, and such experiment will take longer than that in order to activate the frozen cells, pre-grow them and also the fermentation itself that takes more than 6 days, besides analyzing the samples by HPLC that takes its time also!

Comments 7: Finally, the authors discuss the phylogeny of beta-glucosidase without reference or tree for support. In figure 7 the authors propose a phylogenetic tree of cellobiose transporters and I would have expected to root it some well known hexose transporters instead of a putative Trichoderma reesei cellobiose transporter.

Response 7: Our revised manuscript brings now a phylogenetic tree of beta-glucosidases (new Figure 7) and the phylogenetic tree for cellobiose transporters (new Figure 8) includes known hexose and maltose transporters from yeasts. We believe that the figures are better now.

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript deals on the construction of a strain of S. cerevisiae that overexpresses an intracellular βglucosidase (SpBGL7) from Spathaspora passalidarum, and co-expresses the cellobiose transporter SiHXT2.4 from Scheffersomyces illinoinensis, and two putative transporters from Candida tropicalis (CtCBT1) and Meyerozyma guilliermondii (MgCBT2) for the production of second-generation (2G) ethanol from xylose and cellobiose. Authors used the CRISPR-cas-9 system to integrate the SpBGL7 and MgCBT2 genes into S. cerevisiae. The manuscript is on the topics of the journal, but the recombinant strain MP-B7-CBT2DNDC did no let the total simultaneous cellobiose and xylose consumption. Therefore, the overexpressing cellobiose transporters, was not an efficient strategy to consume simultaneously both carbohydrates and produce ethanol 2G. Thus, authors focused the discussion on the importance of the ubiquitinylation of lysine residues at the N- or C-terminal domains of the permease, to explain the results obtained. Therefore, the presented biological system fails as a tool for the production of 2G alcohol.

Comments:

Line 19: In this study, we cloned and expressed in the S. cerevisiae CEN-PK2-1C strain …

Line 22…. (MgCBT2) genes

Line 24: delete stuck and the parenthesis

Line 67: delete “favored”

Line 71: delete ”interesting”

Lines 96 to 99: indicate the complete name of the microorganisms.

Lines 100-111: Do not include your results in the introduction section.

Lines 432-433: re-write the sentence. It is so colloquial

Figure 6 shows that the recombinant strain MP-B7-CBT2DNDC did no let the total simultaneous cellobiose and xylose consumption. Therefore, the overexpressing cellobiose transporters, was not an efficient strategy to consume simultaneously both carbohydrates.

Comments on the Quality of English Language

There are few English style corrections, but the PDF format did not let to make corrections on the file.

Author Response

Comments 1: The manuscript deals on the construction of a strain of S. cerevisiae that overexpresses an intracellular βglucosidase (SpBGL7) from Spathaspora passalidarum, and co-expresses the cellobiose transporter SiHXT2.4 from Scheffersomyces illinoinensis, and two putative transporters from Candida tropicalis (CtCBT1) and Meyerozyma guilliermondii (MgCBT2) for the production of second-generation (2G) ethanol from xylose and cellobiose. Authors used the CRISPR-cas-9 system to integrate the SpBGL7 and MgCBT2 genes into S. cerevisiae. The manuscript is on the topics of the journal, but the recombinant strain MP-B7-CBT2DNDC did no let the total simultaneous cellobiose and xylose consumption. Therefore, the overexpressing cellobiose transporters, was not an efficient strategy to consume simultaneously both carbohydrates and produce ethanol 2G. Thus, authors focused the discussion on the importance of the ubiquitinylation of lysine residues at the N- or C-terminal domains of the permease, to explain the results obtained. Therefore, the presented biological system fails as a tool for the production of 2G alcohol.

Response 1: We would like to thank the reviewers for their comments/suggestions/criticisms, which we feel contributed to improve our manuscript. Detailed answers to the points raised by each reviewer are given below, and all modifications in the text are marked with yellow. Regarding the xylose-cellobiose co-fermentation, the results were a surprise for us and we intend to try to improve the performance of the strains in the future. One thing that we are now highlighting is the cellobiose fermentation data from yeast strains harboring several transporters/enzymes (see new Table 3), showing that indeed cellobiose fermentation is slower than the fermentation of other sugars, with also lower ethanol yields. Thus, improvements are still needed for production of 2G ethanol…..

Comments:

Line 19: In this study, we cloned and expressed in the S. cerevisiae CEN-PK2-1C strain …

Correction done.

Line 22….(MgCBT2) genes

Correction done.

Line 24: delete stuck and the parenthesis

Correction done.

Line 67: delete “favored”

Correction done.

Line 71: delete ”interesting”

Correction done.

Lines 96 to 99: indicate the complete name of the microorganisms.

Corrections done.

Lines 100-111: Do not include your results in the introduction section.

Response: We followed the instructions of the Fermentation Word template, which says regarding the Introduction section: “Finally, briefly mention the main aim of the work and highlight the principal conclusions.”

Lines 432-433: re-write the sentence. It is so colloquial

Correction done. We hope that now is better (now L436-437)

Figure 6 shows that the recombinant strain MP-B7-CBT2DNDC did no let the total simultaneous cellobiose and xylose consumption. Therefore, the overexpressing cellobiose transporters, was not an efficient strategy to consume simultaneously both carbohydrates.

Response: Yes, the cloned and engineered cellobiose transporter failed to consume all cellobiose in xylose-cellobiose co-fermentations, and we intend to address this issue in future work.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript addresses an important and timely topic related to the degradation of lignocellulosic biomass for biofuel production. The subject matter is compelling, and the results are well-documented. The authors have effectively demonstrated the action of the cloned genes, and the experimental methodology is sound and well executed.

However, the manuscript falls short in connecting the results with the broader research objectives outlined in the introduction, particularly the potential for future commercial applications. The experiments demonstrate the effectiveness of the engineered strains at low substrate concentrations and over extended periods. Similar modifications of Saccharomyces cerevisiae strains have been explored for years without achieving the efficiency needed to justify commercial viability. The introduction emphasizes the high costs associated with enzymes used in lignocellulosic biomass hydrolysis, yet the results and discussion sections do not provide sufficient data or analysis to assess whether the observed improvements could eventually contribute to cost-effective and economically viable bioethanol production.

The authors should critically address the broader context of their findings by comparing their engineered strains to other modified yeast strains developed for commercial bioethanol production, many of which have shown promise in laboratory settings but have failed to scale due to insufficient efficiency. A comparative analysis with existing literature on yeast engineering efforts would provide valuable insight into the current limitations and highlight the novel contributions of this study.

It would also be beneficial to compare the performance of the engineered yeast strains with alternative systems where lignocellulosic biomass breakdown and fermentation are conducted in separate steps. This comparison could offer a practical perspective on the potential advantages or disadvantages of using a single genetically modified organism for the entire process versus a multi-organism approach.

Author Response

Comments 1: The manuscript addresses an important and timely topic related to the degradation of lignocellulosic biomass for biofuel production. The subject matter is compelling, and the results are well-documented. The authors have effectively demonstrated the action of the cloned genes, and the experimental methodology is sound and well executed.

Response 1: We would like to thank the reviewers for their comments/suggestions/criticisms, which we feel contributed to improve our manuscript. Detailed answers to the points raised by each reviewer are given below, and all modifications in the text are marked with yellow.

Comments 2: However, the manuscript falls short in connecting the results with the broader research objectives outlined in the introduction, particularly the potential for future commercial applications. The experiments demonstrate the effectiveness of the engineered strains at low substrate concentrations and over extended periods. Similar modifications of Saccharomyces cerevisiae strains have been explored for years without achieving the efficiency needed to justify commercial viability. The introduction emphasizes the high costs associated with enzymes used in lignocellulosic biomass hydrolysis, yet the results and discussion sections do not provide sufficient data or analysis to assess whether the observed improvements could eventually contribute to cost-effective and economically viable bioethanol production.

Response 2: In the Introduction we tried to highlight the rationally for developing cellobiose fermenting yeasts through direct transport and intracellular hydrolysis of the sugar, in the context of xylose-cellobiose co-fermentation. However, our results indicated that the cloned transporter did not allow efficient co-fermentation. Nevertheless, in the new Table 3 we summarize results on cellobiose fermentation by recombinant strains, especially by industrial strains that after selection (evolution) show remarkable yields (0.50 g ethanol/g of cellobiose, 98% of the maximum possible yield) in a very short time (18 h). We have also discussed how truncation of hexose transporters (e.g. HXT1) can improve xylose fermentation, using real industrial substrates (molasses mixed with hemicellulose hydrolyzate) (see L521-525).

Comments 3: The authors should critically address the broader context of their findings by comparing their engineered strains to other modified yeast strains developed for commercial bioethanol production, many of which have shown promise in laboratory settings but have failed to scale due to insufficient efficiency. A comparative analysis with existing literature on yeast engineering efforts would provide valuable insight into the current limitations and highlight the novel contributions of this study.

Response 3: We hope that now the discussion, including the new Table 3, addresses the limitations and potential solutions for improving the fermentation potential of recombinant strains.

Comments 4: It would also be beneficial to compare the performance of the engineered yeast strains with alternative systems where lignocellulosic biomass breakdown and fermentation are conducted in separate steps. This comparison could offer a practical perspective on the potential advantages or disadvantages of using a single genetically modified organism for the entire process versus a multi-organism approach.

Response 4: We agree with the reviewer that several other alternative systems (e.g. SSF, simultaneous saccharification and fermentation) can be employed, including process with a consortium of microorganisms, but we decided to focus in simple batch cellobiose fermentation by recombinant yeasts, with several manuscripts already published. Our current Discussion is already 4 pages long, and we did not want to extend it more…..

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have addressed most of my comment satisfactorily. I noted a slight discrepancy in the number of transporters tested. In the text the authors refer to 4 transporters : SiHXT2.4, 2 CtCBT and MgCBT2 but only 3 are named. Is one of the 2 Ct transporters not tested?

The authors do not discuss the difference in yield of the truncated MgCBT2 which is reversed with the fact that the double truncated MgCBT2 consumes more sugar.

Comments on the Quality of English Language

line 22 maybe replace while with only 3 trasnporters.

lines 106-107 I would state transporter downregulation through endocytosis.

line 419 where instead of were.

line 545 delete more

Author Response

Comments 1: The authors have addressed most of my comment satisfactorily. I noted a slight discrepancy in the number of transporters tested. In the text the authors refer to 4 transporters : SiHXT2.4, 2 CtCBT and MgCBT2 but only 3 are named. Is one of the 2 Ct transporters not tested?

Response 1: We would like to thank again the reviewer for their comments/suggestions, and all new modifications in the text are marked with yellow. We apologize for the confusion, only 3 transporters were tested, and thus we changed that part to “…..SiHXT2.4 from Scheffersomyces illinoinensis, and two putative transporters, one from Candida tropicalis (CtCBT1 gene), and one from Meyerozyma guilliermondii (MgCBT2 gene).”(see L21-22 of the revised manuscript). We hope that now is more clear…… the transporters from C. tropicalis and M. guilliermondii are annotated in their genomes as “putative transporters”.

Comments 2: The authors do not discuss the difference in yield of the truncated MgCBT2 which is reversed with the fact that the double truncated MgCBT2 consumes more sugar.

Response 2: We now included in the Discussion the following: “Although the ethanol yield of the strain with the MgCBT2ΔNΔC transporter was lower than the ethanol yield of the unchanged MgCBT2 permease, there was a clear advantage with the MgCBT2ΔNΔC transporter in terms of ethanol titer, as the transporter truncated in the N- and C-terminal domains produced 33.7% more ethanol than the MgCBT2 permease, and more than doubled the amount of ethanol produced with the MgCBT2ΔC transporter (Figure 5).” (see L533-538 of the revised manuscript).

Comments 3: line 22 maybe replace while with only 3 trasnporters.

Response 3: Given the changes made in L21-22 regarding the number of transporters (see Response 1), we believe that changing “….while with only 3…) is not necessary any more, and thus old L22 was kept unchanged (“While all three transporters allowed.....” now L23 of the revised manuscript).

Comments 4: lines 106-107 I would state transporter downregulation through endocytosis.

Response 4: changed as suggested (L106-107 of the revised manuscript).

Comments 5: line 419 where instead of were.

Response 5: corrected as suggested (L419 of the revised manuscript).

Comments 6: line 545 delete more.

Response 6: deleted as suggested (now L550 of the revised manuscript).

Reviewer 2 Report

Comments and Suggestions for Authors

Main corrections were included, it is a standard paper and the topic is interesting on the biofuel production research. 

Author Response

Comments: Main corrections were included, it is a standard paper and the topic is interesting on the biofuel production research.

Response: We would like to thank again the reviewer for its contributions to improve our manuscript.