eGFP Gene Integration in HO: A Metabolomic Impact?

Round 1

Reviewer 1 Report

In this manuscript Authors presented the data on the Green Fluorescent Protein (GFP) gene integration into HO gene responsible for the diploidization of haploid Saccharomyces cerevisiae cells, and studied the metabolome of the obtained yeast mutant under alkoholic fermentation. Also, phenotype analysis including growth rate and fermentation process were performed. The Authors showed that the GFP integration into HO locus doesn’t have impact on yeast growth rate (as expected based on the published data), yet metabolomic analyses of the wild–type and mutant strain revealed some features discriminated between the strains.

The expression of GFP, as a reporter gene or molecular marker etc., in yeast cells is a common approach therefore the data on its consequences for yeast growth and metabolism are of great deal of general interest. The submitted manuscript is responding to this issue, yet in my opinion more detailes should be given.

• The Authors use ‘eGFP’ within the manuscript (though in the title there is GFP alone). Is it GFP or enhanced GFP (EGFP) that was used? Also, the simple figure ilustrating working model with HO locus, homology arms and GFP (as casettes) would benefit the manuscript.
• My main consideration is about the specificity of the observed phenotype. Reading the data I’m not convinced that all of the observed effects are due to the expression of GFP and not due to deletion of the HO. Many reports showed that the integration of a gene into HO locus doesn’t have impact on yeast growth/phenotype, but how many of them were analysed in terms of metabolome modification? Analysis with MS is sensitive enough, so including a yeast mutant expressing other fluorescent protein fused with Nourseothricin or Nourseothricin alone in HO locus as a control would help for better interpretation of the data, and elevate the value of it.
• The Authors have shown the fluorescence of GFP in yeast mutant. How about showing the level of GFP gene and/or protein expression by qPCR and blotting, respectively?
• The phenotype screening analysis of the yeast mutant(s) exposed to different stress conditions (e.g. availability of diverse carbon sources, osmotic stress, high/low temperature stress, oxidative stress, proteotoxic stress, etc.) is also worth trying.
• The raw data presenting the annotated compounds/ identified biomarkers should be included to the manuscript as supplementary data.
• Minor spell check such as in line 116- as follow(s) or line 211 recombinante, is reqiured.

Author Response

Please see the attachment 

Author Response File: Author Response.pdf

Reviewer 2 Report

In the article entitled “GFP gene integration in HO: a metabolomic impact?” a phenotypic analysis of a commercial S. cerevisiae strain with eGFP integrated in the HO gene is presented. The objective of this publication is clear and relevant. However, we cannot conclude that eGFP integration in the HO gene drives the observed and reported changes since important experimental controls are missing. Moreover, access to all identified metabolites is important, it could be provided as external repository or as a raw-data table. I have the following questions:

1. What is the phenotype, at fermentation and exometabolome level, in a cerevisiae S3 strain with HO gene disrupted but without eGFP integration?
2. What is the phenotype, at fermentation and exometabolome level, in a cerevisiae S3 strain with eGFP integrated in non-coding region in the genome?

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

In this paper, the authors integrated an eGFP expression cassette into the HO locus of a commercial yeast strain and investigated its impact by fluorescence microscopy, growth measurements, and "exometabolome" analysis. Integration of the eGFP cassette into the HO locus had no significant effect on growth or fermentation capacity, but exometabolome analysis identified molecules that differed in abundance with and without eGFP expression.  The authors claim that exometabolome analysis can be used to scrutinize differences in the properties of recombinant yeast.

The experiments were performed in a scientifically correct manner, and the conclusions reached are not unreasonable. However, there are some concepts that are missing from the background of this study, and the presentation of the data is inadequate and needs to be revisited. The specifics are discussed below.

Major Points

1) The effect of (massive) expression of eGFP
The authors focus only on the effects of integration of the eGFP cassette into the HO locus, which occurs when the HO locus is disrupted. However, the expression of eGFP, or gratuitous proteins in general, can negatively affect cell growth. There is no mention of this phenomenon, which is called protein loading or protein cost (ex. PMID26725116, PMID: 30095406, PMID: 26543202). The authors express eGFP using the TEF2 promoter, which is one of the most powerful promoters, and the expression of gratuitous proteins like eGFP from this promoter can overload the yeast protein synthesis system.

Although no growth inhibition was observed in the experiments conducted by the authors, it is possible that effects could be detected in sensitive competitive growth experiments as performed in PMID26725116. I am not suggesting that new experiments are needed, but this concept or concern should be mentioned in the introduction or discussion.

Related to this, the possibility of eGFP-derived peptides being detected by exometabolome analysis should also be considered. For example, the peptides obtained when purified eGFP is analyzed by the authors' method may or may not match the peptide markers obtained in eGFP-expressing strains.

(2) Statistical Analysis of Exometabolome
The authors conclude that the abundance of 285 (10%) of the molecules detected was "significantly" different between eGFP-expressing and control strains.  For this detection, a t-test of p < 0.01 was used as the significance level. However, the analysis conducted by the authors was a multiple test, and the significance level could be much lower.The authors should perform an FDR control for multiple testing and come up with a significance level at the q value. Then the number of molecules whose abundance "significantly" changed would be much smaller.

Minor points

1) Line 103: What is the unit "UDO"?  Spell out and explain.

2) Figure 1: The colors are very difficult to identify. Create a series of pictures with the colors divided by channel.  Change the color of the scale bar. It is not legible. Describe what the blue and green channels represent in the legend, and mention whether the exposure time is the same or different between the images.

3) Line 228: "This signal seemed to be emitted at the region of the cell wall..." In Figure 1A, the blue color is visible inside the cell. Do we see different things in Figure 1A and 1B?  What is the reason for this? Again, create images for each channel and properly explain the differences between them.

4) Line 229: "... a significant difference in green fluorescence ...", Line 231: "... less intense than for the modified strain.” How was the fluorescence quantified, and how was "a significant difference" obtained by statistical processing?

5) Figure 2: Vertical axis "Cell viability" may be an error for "Viable cell" Also, how was this calculated? I assume it is the flow cytometry analysis described in Materials and Methods, but the authors only describe staining (of dead cells) with PI. Please describe how you detected "living" cells. Also, italicize "Saccharomyces cerevisiae" in the legend.

(6) Lines 277-8: Also mention the effect of eGFP expression as described above.

7) Figure 4: No color index in A. Shouldn't "Amino acids" in B1 and B2 be "Amino acids or Peptides"?

8) Lines 307-8: "We observed a low diversity of composition between the biomarkers of the two strains." Please elaborate on what part of Figure 4 leads you to such a conclusion.

(9) Lines 284-5: As mentioned above, examine the significance level in consideration of multiple testing.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The article: “eGFP gene integration in HO: a metabolomic impact?” has been corrected and the Authors have adressed my comments.

Previously, my main consideration was about the specificity of the observed phenotype of the mutant strain expressing EGFP. The Authors haven’t performed any additional experiments suggested, however they explained clearly the reasons both in the manuscript and coverletter and I respect the arguments mentioned.

Yet, I still have some minor issues:

In the corrected version the numbers of features identified by MS have been changed. It doesn’t affect the conclusions made, yet the question raised why the results were changed at this stage?

In the Supplemental Fig.1. there are turquise and navy blue parts present after the Cas9 endonuclease cleavage. What are they representing? I think they should be removed.

In the Introduction the Authors claim that Line 33 :“ The Green Fluorescent Protein (EGFP), discovered in 1962 by Shimomura 33 [6] and derived from the jellyfish Aequorea vitoria”- this sentence should be corrected as GFP (not EGFP) was discovered by Shimomura in 1962. Also, in A. vitoria there is a letter missing (victoria) and in line 58 there is repetition of EGFP.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors have addressed my concerns. 

Author Response

Thank you for your comments