Placental Galectin-2 Expression in Gestational Diabetes: A Systematic, Histological Analysis

Round 1

Reviewer 1 Report

Hepp et al, provide a detailed analysis of Galectin-1 expression in placental tissue of control and gestational diabetes (GDM) subjects. From their analysis they conclude that Galectin-1 expression is significantly high in the GDM placenta irrespective of the fetus gender. While this study is relevant to expanding our current understanding of processes associated with GDM, parts of the study require clarification:

1. Are there any distinction in the placental structures of control and GDM subjects?
2. Representative images of Galectin-1 staining in control and GDM in syncytiotrophoblast shown in Figure 2 indicate that this particular tissue is more disintegrated in the GDM? If true, the authors should discuss this observation in the manuscript. If this is incorrect, then the authors should provide an alternative representative image.
3. For clarity, the “maternal” decidua and the “fetal” syncytiotrophoblast should be written in the manuscript.
4. The authors show that GDM placenta expresses higher Galectin-1, but these subjects also have BMI skewing higher than the control. Is there a possibility that subjects with higher BMI may have greater Galectin-1 expression? Analyzing Galecitn-1 expression in appropriate BMI control should be included for clarity. The authors should also provide a breakdown of number of control and GDM subjects in the indicated BMI range.
5. Since there is already a study implicating Galectin-1 in GDM, the authors should refrain from using the word “first time” while describing their findings.
6. Given the studies mostly anti-inflammatory function of Galectin-1, the increased expression of Galectin-1 maybe an end result of inflammation induced by GDM rather than a potential mechanism causing GDM. Can the authors comment on this possibility?
7. On the same note, if the Galectin-1 is proinflammatory, the authors should consider staining the maternal decidua tissue from control and GDM subjects to further support the role of Galectin-1 in GDM and increased inflammation.
8. The authors show dynamic plots (bar graphs) of the data, these plots are not recommended in scientific literature as they conceal underlying data trends and may also be misguiding. Please see 'https://simplystatistics.org/2019/02/21/dynamite-plots-must-die/' for issues and alternatives for dynamic plots.
9. I strongly recommend the manuscript to be reviewed by a native English speaker. Here are just a few examples of issues concerning scientific writing:
   1. In Fig 1 legend on page 3, line 26, the authors state, "There was no statistically significant between male ..." I believe the correct version of this needs the missing word 'difference' so the correct version would read, "There was no statistically significant difference between male ...."
   2. Section 2.1 describes figure 1. But the definition of abbreviation for "IRS" is defined in section 2.2. Readers who are not subject matter experts will find it challenging to follow the manuscript for such instances. I encourage the authors to re-write such cases if any more exists.
10. The immunohistochemical stainings are shown with 100-fold and 400-fold zoomed versions. It seems like the 400-fold version is a part of 100 fold versions. If this is true, I recommend the authors to add annotations (a little square box highlighting the regions represented in 400-fold staining).

Author Response

Response to the Reviewers

“Galectin-2 expression is upregulated in Gestational Diabetes”

Dear Ms Xie, Dear Reviewers,

We kindly thank you for your detailed and thorough evaluation of our manuscript. Your comments, correction and questions were highly appreciated and of great help for further improving the quality of said manuscript. In the process of revising the paper, we were careful to take all your considerations into account and prepared the following point to point response.

Sincerely,

The Authors

Reviewer #1

Hepp et al, provide a detailed analysis of Galectin-1 expression in placental tissue of control and gestational diabetes (GDM) subjects. From their analysis they conclude that Galectin-1 expression is significantly high in the GDM placenta irrespective of the fetus gender. While this study is relevant to expanding our current understanding of processes associated with GDM, parts of the study require clarification:

In general, we investigated gal-2 in GDM

1. Are there any distinction in the placental structures of control and GDM subjects?

Overall no significant distinctions could be observed between GDM and control placentas in terms of morphology.

2. Representative images of Galectin-1 staining in control and GDM in syncytiotrophoblast shown in Figure 2 indicate that this particular tissue is more disintegrated in the GDM? If true, the authors should discuss this observation in the manuscript. If this is incorrect, then the authors should provide an alternative representative image.

Taking into consideration the previous comment, we have changed the image in Fig. 2 to a more appropriate one.

3. For clarity, the “maternal” decidua and the “fetal” syncytiotrophoblast should be written in the manuscript.

We added “maternal” and “fetal” throughout the main text of the manuscript (e.g. see page 3, lines 14&15).

4. The authors show that GDM placenta expresses higher Galectin-1, but these subjects also have BMI skewing higher than the control. Is there a possibility that subjects with higher BMI may have greater Galectin-1 expression? Analyzing Galecitn-1 expression in appropriate BMI control should be included for clarity. The authors should also provide a breakdown of number of control and GDM subjects in the indicated BMI range.

This is of course a valid concern. We included statistical testing of galectin-2 expression in patients with low to normal BMI (<25kg/m2) compared to overweight and obese patients (≥25kg/m2) [1]. There was no significant difference between the two groups. We edited the result (see newly added section 2.5.) as well as the M&M section (see Table 3, page 8) to include these findings, as well as the requested data on control vs. GDM by BMI.

5. Since there is already a study implicating Galectin-1 in GDM, the authors should refrain from using the word “first time” while describing their findings.

Thoroughly searching the PubMed library using the terms “galectin-2” with “gestational diabetes”, “GDM”, “diabetes” “placenta” and “gestation” we were unable to find any previous of work on galectin-2 expression in GDM. Perhaps there was a slight mix-up concerning the different galectins previously described in GDM, because you did mention gal-1 in the review? However, if you prefer for us to leave out the term “first time” please inform us and we will be happy to do so in the final version.

6. Given the studies mostly anti-inflammatory function of Galectin-1, the increased expression of Galectin-1 maybe an end result of inflammation induced by GDM rather than a potential mechanism causing GDM. Can the authors comment on this possibility?

In their paper from 2014 Blois, et al. [2] were able to show a significantly higher expression of galectin-1 in GDM placentas. Within the discussion the hypothesis was put forward – as you also suggest –, that the increased expression may result from the inflammatory state of GDM rather than causing it. Having now investigated the expression of galectin-2 in GDM, we show an increased placental expression in GDM cases. Both galectin-1 and galectin-2 show anti-inflammatory function. Based on Blois, et al. [2] and other studies, we thus discussed the possibility that the elevation of galectin-2 in GDM placentas may be an end result of the disease rather than the cause as one option (see page 6, lines 17-21). We have modified our discussion to make the argument sufficiently clear to the reader (see page 7, lines 1-3 and concluding paragraph).

7. On the same note, if the Galectin-1 is proinflammatory, the authors should consider staining the maternal decidua tissue from control and GDM subjects to further support the role of Galectin-1 in GDM and increased inflammation.

Results of immunohistochemical staining of maternal decidua tissue have been included in the manuscript (please see section 2.3., page 4). The respective figure is also shown below:

8. The authors show dynamic plots (bar graphs) of the data, these plots are not recommended in scientific literature as they conceal underlying data trends and may also be misguiding. Please see 'https://simplystatistics.org/2019/02/21/dynamite-plots-must-die/' for issues and alternatives for dynamic plots.

Thank you for pointing this out! We have modified our graphs according to “simplystatistics.org’s” suggestion. In order to maintain somewhat of a graphical summary of the IRS results, we kept the box plot and chose to use on overlay technique (see Figures 1-4).

9. I strongly recommend the manuscript to be reviewed by a native English speaker. Here are just a few examples of issues concerning scientific writing:

* In Fig 1 legend on page 3, line 26, the authors state, "There was no statistically significant between male ..." I believe the correct version of this needs the missing word 'difference' so the correct version would read, "There was no statistically significant difference between male ...."

* Section 2.1 describes figure 1. But the definition of abbreviation for "IRS" is defined in section 2.2. Readers who are not subject matter experts will find it challenging to follow the manuscript for such instances. I encourage the authors to re-write such cases if any more exists.

Thank you for pointing this out to us. We have conducted a thorough review concerning scientific language and corrected (among others) the errors mentioned above.

10. The immunohistochemical stainings are shown with 100-fold and 400-fold zoomed versions. It seems like the 400-fold version is a part of 100 fold versions. If this is true, I recommend the authors to add annotations (a little square box highlighting the regions represented in 400-fold staining).

Yes, the 400-fold version is contained within in the 100-fold. We therefore followed your suggestion and indicated which exact region is represented in the inserted zoom (see Fig. 2 and 3).

Author Response File: Author Response.pdf

Reviewer 2 Report

In this manuscript, Hepp et al. analyse Galectin-2 expression in gestational diabetic mellitus (GDM) placentas. Their finding is that it is upregulated using a semi-quantitative method. The manuscript can be (and should) greatly improved before its publication. This indication is based on the comments below.
Specific comments: (mainly in the order shown in the ms, not in order of importance)
- There is known information about Galectin-2 related to insulin and inflammation which is not mentioned in the Introduction. For an example see Brinchmann et al. review (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5987346/).
Page2:
- -Line 14:”…many mechanism…” should be “…many mechanisms…”.
- -Line 41: comma after “Remarkably…” missing.
- Line 42: the cardiovascular risk mentioned, is it in the progeny of GDM patients? Please specify.
- Lines 32, 38 and 50: space missing before the reference number.
- On occasions, the authors write Galectin-2 while in other places, galectin-2 is written. Please homogenise the nomenclature.
- Line 48 and beginning of 49: the sentence does not make sense: “Galectin-2 belongs to the prototype galectins, constituting that is contains only one CRD;…”.
Page 3:
- Figure 1: what are the numbers next to the circles? Please specify in the figure legend. IRS abbreviation (line 22) is used before indicating its meaning (line 31).
- On the IRS method:
o Why is the median IRS value given instead of the average?
o How many slides per patient were measured?
o How many times was the staining performed?
o How many microscopic fields were counted per sample?
o How many cells were counted to determine the percentage of expressing cells?
o How many people did the counting and on how many sessions? Both things can affect subjectively the scoring.
Page 4:
- Scale bars in pictures and high magnification images cannot be read at that size.
- No need to repeat in every figure legend (unless required by the journal specs) the meaning of the box plots.
- Rather than mentioning the magnification, the authors should indicate the objectives used. Also, rather than giving this information in the figure legend, it should be included in the M&M.
- Line 24: ”…triple filter excitation microscopy…” is an odd name for fluorescent microscopy. Also, the authors should revise the grammar in that sentence.
- Lines 26-27: the authors should include the references were it was shown that CK7 and CD32 are markers for EVTs and foetal endothelial cells, respectively.
Page 5:

Immunofluorescence analysis shown in Figs. 4-5:
o The magnification shown is rather low and it’s difficult to see CD31 or CK7 staining.
o Arrows pointing to colocalizing staining would help the reader to follow the authors indications to colocalizing staining.
o To homogenise the figure legends, all used fluorophores should be indicated, not just for galectin-2. The authors should indicate which blue staining was used for the nuclei: DAPI or Hoechst. Also the green fluorophore should be specified.
o Nuclear channel alone should also be shown: galectin-2 localization seems to be cytoplasmic in Fig. 4 (both control and GDM) and in GDM in 5. However, control in Fig. 5 looks like nuclear. Is that correct? If so, could the authors explain why this might be? Alternatively, the colocalizing spots in controls in Fig. 5 might represent some kind of debris with autofluorescence in the 3 channels.
o Why do the cells in controls in Fig. 4 look smaller than those in GDM? Do the authors have an explanation for this?
- Line 7: “…identifying EVTs as the PREDOMINANT galectin-2 expressing cell type…”. Could the authors make some kind of quantification on the number of cells expressing galectin-2 in controls and GDM? This would really give definitive information about increased number of cells expressing this marker.
- Related to the previous comment: the authors title includes “UPREGULATION”. The authors did not perform any experiment to measure galectin-2 expression (either RNA or protein level). The authors should consider changing the title of the manuscript as it is misleading.
Page 6:
- Line 32-33: are the authors implying that placentas secrete insulin? Maybe this reviewer is missing something important here. The authors should rewrite this paragraph taking into account that they haven´t shown upregulation in galectin-2 expression and it is not known whether in DM there is higher expression of galectin-2.
- Line 45: “…however…” should be between commas.
Page 7:
- Line 3: delete the “:” at the end of the sentence.
- Line8: comma after “More broadly…” missing.
- Line 16: comma after “In conclusion…” missing.
Page 8:
- Could the authors please indicate where does the blue staining in Fig. 8B come from? What is the nature of the so called “negative-control-serum”?

Author Response

Reviewer #2

In this manuscript, Hepp et al. analyse Galectin-2 expression in gestational diabetic mellitus (GDM) placentas. Their finding is that it is upregulated using a semi-quantitative method. The manuscript can be (and should) greatly improved before its publication. This indication is based on the comments below.

Specific comments: (mainly in the order shown in the ms, not in order of importance)

Page1:

- There is known information about Galectin-2 related to insulin and inflammation which is not mentioned in the Introduction. For an example see Brinchmann et al. review (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5987346/).

We have reviewed the suggested review as well as the PubMed library once more. An additional paragraph was added to the introduction, highlighting the current knowledge on galectin-2 in relation to insulin resistance and cardiovascular disease (see page 3, lines 1-7).

Page2:

-Line 14:”…many mechanism…” should be “…many mechanisms…”.

-Line 41: comma after “Remarkably…” missing.

Thank you for your many, highly attentive alerts concerning grammar and spelling. We carefully corrected them and re-read the manuscript again for further errors.

- Line 42: the cardiovascular risk mentioned, is it in the progeny of GDM patients? Please specify.

The cardiovascular risk mentioned at this point refers to the general population, as well as GDM positive mothers. We have adjusted the text in order to make this clear to the reader (see page 2, lines 41-42).

- Lines 32, 38 and 50: space missing before the reference number.

Spaces have been added before all reference numbers.

- On occasions, the authors write Galectin-2 while in other places, galectin-2 is written. Please homogenise the nomenclature.

The nomenclature has been cleared up; minuscular is now used throughout the text.

- Line 48 and beginning of 49: the sentence does not make sense: “Galectin-2 belongs to the prototype galectins, constituting that is contains only one CRD;…”.

The sentence was rephrased (see page 2, lines 48-49).

Page 3:

- Figure 1: what are the numbers next to the circles? Please specify in the figure legend. IRS abbreviation (line 22) is used before indicating its meaning (line 31).

The numbers in the circles indicated which sample number(s) belonged the outlier(s). However, the graphs have been changed and now include a scatter of all data points, making the sample number of outliers irrelevant. Thus, in the current version it has been deleted.

The meaning of IRS abbreviation is now indicated in line 22.

- On the IRS method:

We have added a more detail to our M&M section in order to clarify the queries blow for the reader (see page 9, lines 22-26).

• Why is the median IRS value given instead of the average?

Due to the fact that IRS values are semi-quantitative hence not metric but rather ordinal in terms of their statistical characteristic the average is not applicable here.

• How many slides per patient were measured?

One slide per patient was measured

• How many times was the staining performed?

Staining was performed once, unless the quality of the slide or staining was unsatisfactory in which case it was repeated until we obtained a functional slide.

• How many microscopic fields were counted per sample? How many cells were counted to determine the percentage of expressing cells?

3-5 microscopic fields with a minimum of 100 cells were counted per sample.

• How many people did the counting and on how many sessions? Both things can affect subjectively the scoring.

All slides were evaluated by two independent observers, with no more than two sessions per observer. In cases where the observers came to different conclusions, the sample was re-evaluated and discussed until reaching one conclusive result.

Page 4:

- Scale bars in pictures and high magnification images cannot be read at that size.

We increased the size of Figures 2 and 3. Hopefully all aspects are now of satisfactory size and quality.

- No need to repeat in every figure legend (unless required by the journal specs) the meaning of the box plots.

We modified the figure legends now referring to Fig. 1 for box plot explanation.

- Rather than mentioning the magnification, the authors should indicate the objectives used. Also, rather than giving this information in the figure legend, it should be included in the M&M.

We adjusted the manuscript accordingly. The M&M section now includes objectives used. We also applied this suggestion to the section on double immunofluorescence.

- Line 24: ”…triple filter excitation microscopy…” is an odd name for fluorescent microscopy. Also, the authors should revise the grammar in that sentence.

We have modified this sentence, adjusting the vocabulary as well.

- Lines 26-27: the authors should include the references where it was shown that CK7 and CD32 are markers for EVTs and foetal endothelial cells, respectively.

The respective references have been added (see page 5, line 8 and references 58 and 59).

Page 5:

Immunofluorescence analysis shown in Figs. 4-5:

• The magnification shown is rather low and it’s difficult to see CD31 or CK7 staining.

All pictures were taken through a 63-fold microscopic objective. While some additional pictures were taken through a 40-fold objective, no magnification higher than 63-fold was used. Due to the unfortunate fact, that the fluorescence period of the slides is limited at this point we are unable to include new pictures at higher magnifications. However, we were able to include cropped and zoomed version of the displayed pictures as a supplementary figure. The figure legend indicates, that the pictures were generated using a photo editing software.

• Arrows pointing to colocalizing staining would help the reader to follow the authors indications to colocalizing staining.

The requested arrows were added to figures 4 and 5.

• To homogenize the figure legends, all used fluorophores should be indicated, not just for galectin-2. The authors should indicate which blue staining was used for the nuclei: DAPI or Hoechst. Also, the green fluorophore should be specified.

We have standardized the respective figure legends and included the missing information.

• Nuclear channel alone should also be shown: galectin-2 localization seems to be cytoplasmic in Fig. 4 (both control and GDM) and in GDM in 5. However, control in Fig. 5 looks like nuclear. Is that correct? If so, could the authors explain why this might be? Alternatively, the colocalizing spots in controls in Fig. 5 might represent some kind of debris with autofluorescence in the 3 channels.

Depiction of nuclear channel was added Fig. 4 and 5, which we expect should improve the understanding for the reader. The impression of nuclear expression in Fig. 5 seems to stem from an overlay. This can be differentiated when comparing galectin-2 channel with the nuclear channel (see pictures below).

• Why do the cells in controls in Fig. 4 look smaller than those in GDM? Do the authors have an explanation for this?

This seems to be a sort of optical illusion. We have double checked the respective objectives used to generate the images; all were taken through a 63-fold objective. We then proceeded to measuring the respective cells. They did in fact not differ in size (see. pictures below)

- Line 7: “…identifying EVTs as the PREDOMINANT galectin-2 expressing cell type…”. Could the authors make some kind of quantification on the number of cells expressing galectin-2 in controls and GDM? This would really give definitive information about increased number of cells expressing this marker.

Quantification of the phenotypes of galectin-2 expressing cells was provided. The results are displayed in the newly added Table 1 (Page 5). Details were also added to the M&M section.

- Related to the previous comment: the authors title includes “UPREGULATION”. The authors did not perform any experiment to measure galectin-2 expression (either RNA or protein level). The authors should consider changing the title of the manuscript as it is misleading.

The title was changed in order to avoid any misunderstandings. It now reads: “Placental galectin-2 expression in Gestational Diabetes: a systematic, histological analysis”.

Page 6:

- Line 32-33: are the authors implying that placentas secrete insulin? Maybe this reviewer is missing something important here. The authors should rewrite this paragraph taking into account that they haven´t shown upregulation in galectin-2 expression and it is not known whether in DM there is higher expression of galectin-2.

Thank you for calling attention to the misleading phrasing of this paragraph. It was rewritten, taking into consideration the various uncertainties still unclarified concerning galectin-2 (see page 7, lines 4-9).

- Line 45: “…however…” should be between commas.

      Commas were added.

Page 7:

- Line 3: delete the “:” at the end of the sentence.

- Line8: comma after “More broadly…” missing.

- Line 16: comma after “In conclusion…” missing.

The errors in punctuation have been corrected.

Page 8:

- Could the authors please indicate where does the blue staining in Fig. 8B come from? What is the nature of the so called “negative-control-serum”?

As in the participants’ slides, blue staining comes from counterstaining with Mayer´s hemalum. The negative-control-serum contains anti-rabbit-Ig which should not bind to any human epitope, hence serve as negative-controls. We added more detail concerning control staining to the M&M section in order to clarify this (see page 9, lines 6-10).

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors addressed most of my concerns. Overall, the manuscript improved substantially and I think it is now suitable for publication. 

Author Response

Reviewer #1

The authors addressed most of my concerns. Overall, the manuscript improved substantially and I think it is now suitable for publication.

            Thank you, your suggestions from round 1 helped us greatly.

Response to the Academic Editor Notes

While the authors have made all necessary revisions to the manuscript, significant English grammatical errors continue to appear throughout the manuscript and need to be corrected before the manuscript is acceptable for publication in IJMS.

1. For example, the abstract is poorly written.

Page 1, line 19: delete the word “sever”. Should read: “…and negative impacts on the health of both mothers and their offspring over the long-term.”

Page 1, line 19-20 should read: “The molecular mechanisms involved are not fully understood.”

Page 1, line 20-21 should read: “As in other states of insulin resistance, a disproportionate immune response in GDM leads to a state of chronic low-grade inflammation.”

Page 1, line 22 should read: “Galectin-2 exerts regulatory effects on different immune cells.”

Page 1, line 29 should read: “These findings…”

Thank you for these tremendously helpful editorial comments. We integrated them into the abstract and made a few other minor adjustments.

2. Edit the introduction.

Page 2, line 4 delete: “From this staring point…”

Page 2, line 6: “…the majority are expressed…”

Page 2, line 35: GDM is already defined.

Page 3, line 1 change to: “…there are no reports on the role of galectin-2 in GDM…”

We deleted the unnecessary phrases and incorporated the requested adjustments.

3. Edit the discussion.

Page 6, line 14: change the term “significant overexpression” to “increased expression”

Page 6, line 29: change “:” to “.”

Page 6, line 36 change to: “phosphatidylserine”

Page 7, line 2: change the term “over-expression” to “increased expression”

Page 7, line 22: change “:” to “.”

Page 7, line 32: change “the causal” to “a causal”

Page 7, line 44: change the term “galectin-2 overexpression” to “increased galectin-2 expression”

Punctuation and spelling errors have been corrected. Phrasing has been adjusted accordingly.