Effect of Neutrophil–Platelet Interactions on Cytokine-Modulated Expression of Neutrophil CD11b/CD18 (Mac-1) Integrin Complex and CCR5 Chemokine Receptor in Stable Coronary Artery Disease: A Sub-Study of SMARTool H2020 European Project

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have study in a subgroup of patients with coronary artery disease the neutrophils phenotype. 

 and proinflammatory cytokines. Several associations with neutrophils markers were detected.

The inflammatory pathway is very important in chronic CAD.There are several major concerns:

A. The inclusion criteria, legal and ethics aproval, Helsink Declaration, samples collection has to be explained at the begining of materials and methods.

B.Statistical analysis has to be review. Normality of data.

Multivariate regression with 55 included patients has a limitation of variables number including as independent variables.

C.Table with clinical characteristics has to be in results. Risk factors have to be included.

D. Flow cytometry diagram could help to understand the viability of the results 

E. Which is the percentage of positive CD11b/CD18? This has to be explainned.

F. Univaritate regreession analysis have to be shown.

G. Multivariate has to include few number of variables because the sample size is too low.

H. Is there any association with BMI, age or diabetes presence?

The analysis of results have to be improved.

Author Response

Comment 1: The inclusion criteria, legal and ethics approval, Helsinki Declaration, samples collection has to be explained at the beginning of materials and methods.

Response 1: The inclusion criteria have been indicated in the attached Table S1 (Supplementary Material). Regulatory information concerning sample collection, ethics approval, informed consent and Helsinki's Declaration, have been included in the text.

-----------

Comment 2: Flow cytometry diagram could help to understand the viability of the results.

Response 2: A flow cytometry figure (Figure 1) with the corresponding caption has been included in the text (Flow cytometry analysis section).

---------

Comment 3: Which is the percentage of positive CD11b/CD18? This has to be explained.

Response 3: The observed value of percentages of positivity of the studied neutrophil markers have been inserted at the beginning of the Results section.

---------

Comment 4: Is there any association with BMI, age or diabetes presence?

Response 4: The observation that there is no significant association between the studied neutrophil markers (as RFI) and the presence of diabetic disease, as well as with gender distinction, has been included in Results section, immediately after Table 2.

---------

Comment 5: Statistical analysis has to be review. Normality of data.

Response 5: The observation that the non-normal distribution of cytokine data has been corrected by using their log10-based transformation has been inserted at the beginning of Statistical section. Statistical analysis has been changed and reviewed.

--------

Comment 6: Table with clinical characteristics has to be in results. Risk factors have to be included.

Response 6: A table (Table 1) showing several clinical, biochemical, therapeutic and immuno-metabolic parameters of the selected group of patients, has been moved to the Results section. Furthermore, some cardiovascular risk factors (in particular the FRS and creatinine) have been included in the univariate statistical analysis.

--------

Comment 7: Multivariate has to include few numbers of variables because the sample size is too low.

Response 7: As required, the multivariate analysis has been based on a smaller number of adjustment parameters, previously selected by univariate analysis, due to the small size of the patient group.

--------

Comment 8: Univariate regression analysis have to be shown.

Response 8: Univariate regression analysis between cytokine/chemokine data, immuno-biochemical and metabolic parameters able to influence the expression level of neutrophil markers has been shown in Table 2 in the Results section.

---------

Reviewer 2 Report

Comments and Suggestions for Authors

Hello,

Inflammation is an important marker and predictor of CV events and remains an important unconquered territory in the management of CAD.

The study dwells into the "Effect of neutrophil-platelet interaction on cytokine-modulated expression of neutrophil CD11b/CD18 (Mac-1) integrin complex and CCR5 chemokine receptor in stable Coronary Artery Disease as a sub-study of SMARTool H2020 European project."

The SMARTool project is important in addressing the same.

It is a well drafted paper however, A brief background on what is already known about the clinical implications of neutrophil-platelet interaction with regards to clinical events and how GpIIb/IIIa inhibitors have already been used in the clinical practice to address the same may add value to the paper.

Also, the statistical method may be further elucidated in the method section as currently the same can only be inferred partly from the tables and partly form the text. 

Otherwise it is a well drafted paper and can be accepted after minor revision. Thanks

Author Response

Comment 1: It is a well drafted paper however, A brief background on what is already known about the clinical implications of neutrophil-platelet interaction with regards to clinical events and how GpIIb/IIIa inhibitors have already been used in the clinical practice to address the same may add value to the paper.

Response 1: As requested by the Reviewer, and on the basis of our previous experimental results, we have included in the Introduction section a brief comment relating to the clinical utility of the use of inhibitors of the CD41/CD61 platelet receptor complex to control the evolution of atherosclerotic coronary disease.

----------

Comment 2: Also, the statistical method may be further elucidated in the method section as currently the same can only be inferred partly from the tables and partly form the text. 

Response 2: The statistical analysis method has been modified by including a preliminary univariate regression. Its modifications have been described in the Statistical analysis section.

---------

Reviewer 3 Report

Comments and Suggestions for Authors

Reviewer's comments.

Based on what data do the authors write about «cytokine-mediated expression» of adhesion molecules and CCR5?

What was the rationale for studying the particular CCR5 receptor? Is it expressed by neutrophils? Or by platelets that form complexes with neutrophils?

Methods. In what biomaterial were cytokine levels determined? Serum? Plasma?  This is a crucial point, as some cytokines (including chemokines) are released from platelets and leukocytes during blood clotting.

The group of patients is virtually uncharacterized. What parameters confirm the presence of CАD? Therapy?

Why the emphasis on albumin? The authors write that they analyzed lipid profile and plasma adhesion molecules, but do not give the corresponding data.

Manufacturers of antibodies for cytometry, ELISA kits, etc. are not listed.

The description of the cytometry method is too brief. What kind of blood was used (anticoagulant?) - this is a crucial point, as we are talking about the evaluation of leucocyte-platelet aggregates. How long were the samples stored before staining?  What antibodies were used? Was there any additional fixation? etc. The authors should realize that complexes can form/disassemble ex vivo/in vitro.

Results. In regression analysis, coefficients with + and - symbols were obtained. In my opinion and experience, the cytokines IL-6 and TNF cannot "act" differently. The same applies to IL-6 and INFgamma, etc. The authors try to explain this phenomenon in the Discussion section, but it is most likely due to a methodological error rather than to perturbations in the expression of membrane molecules.

In the Discussion section, the authors write about "post-activation-induced apoptotic shedding" of leukocyte beta2-integrins.  Is shedding shown for these molecules? It can hardly be a physiological process.

Author Response

Comment 1:  The group of patients is virtually uncharacterized. What parameters confirm the presence of CАD? Therapy?

Response 1: The parameters confirming the presence of CAD in the group of patients are mainly based on coronary angiography imaging characteristics and described in the cited papers (References 6 and 7). Clinical, immuno-biochemical , therapeutic and metabolic parameters of the selected population are reported in Table 1. Furthermore, the attached Table S1 (Supplementary materials) elucidates further the criteria used for the patients selection.

--------

Comment 2: Manufacturers of antibodies for cytometry, ELISA kits, etc. are not listed.

Response 2: Manufacturers and catalog number codes of antibodies and ELISA kits, as well as of antibodies used for flow cytometric analysis, have been included in the Biochemical analysis and Flow Cytometry analysis sections.

---------

Comment 3: The description of the cytometry method is too brief. What kind of blood was used (anticoagulant?) - this is a crucial point, as we are talking about the evaluation of leucocyte-platelet aggregates. How long were the samples stored before staining?  What antibodies were used? Was there any additional fixation? etc. The authors should realize that complexes can form/disassemble ex vivo/in vitro.

Response 3: The description of flow cytometry analysis has been expanded, including a Figure 1 illustrating the procedure of flow cytometric analysis. Furthermore, informations concerning the anticoagulant used, the time elapsed since sampling and the possible use of fixation procedures, have been provided (Flow cytometry analysis section).

--------

Comment 4: Why the emphasis on albumin? The authors write that they analyzed lipid profile and plasma adhesion molecules, but do not give the corresponding data.

Response 4: The inclusion of plasma albumin levels into the multiple regression analysis models has been motivated by the knowledge, described in the literature, of the important modulatory effect, either pro- or anti-inflammatory, carried out by this protein on cytokine function at a systemic level. To this end, and in particular regarding the stable up-regulation of neutrophil CD11b in the presence of albumin, two References [19,20] have been included in the Discussion section. The quantification of some metabolic, clinical, metabolic parameters and soluble adhesion molecules has been included in Table 1 in the Patients section.

---------

Comment 5: Results. In regression analysis, coefficients with + and - symbols were obtained. In my opinion and experience, the cytokines IL-6 and TNF cannot "act" differently. The same applies to IL-6 and INFgamma, etc. The authors try to explain this phenomenon in the Discussion section, but it is most likely due to a methodological error rather than to perturbations in the expression of membrane molecules.

Response 5: The statistical values of the univariate analysis (Table 2) allow the exclusion of TNF-alfa and IFN-gamma from the subsequent multivariate analysis models. Therefore, any comment relating to the different activating action on neutrophils by IL-6, TNF-alfa and IFN-gamma, as well as to any possible mechanism of shedding of neutrophil surface molecules induced by these cytokines has been eliminated from the Discussion section.

--------

Comment 6: In the Discussion section, the authors write about "post-activation-induced apoptotic shedding" of leukocyte beta2-integrins.  Is shedding shown for these molecules? It can hardly be a physiological process.

Response 6: See the reply to the previous comment.

-------

Comment 7: Based on what data do the authors write about «cytokine-mediated expression» of adhesion molecules and CCR5?

Response 7: We agree with the reviewer in considering the expression of neutrophilic integrin molecules and CCR5 as "modulated" by circulating cytokines and "non-induced" by them. Throughout the paper this concept should be clearly evident.

----------

Comment 8: What was the rationale for studying the particular CCR5 receptor? Is it expressed by neutrophils? Or by platelets that form complexes with neutrophils?

Response 8: In the Introduction section the rationale for studying the modulation of the expression of the CCR5 molecule on neutrophils has been described.

---------

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have addressed the reviewer´s comments. 

Author Response

The authors thank reviewer's comments.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have modified their article somewhat, but the changes made were not satisfactory to me. 

The authors have added the following fragment to the introduction. “In particular, we decided to evaluate the neutrophil expression of the chemokine receptor CCR5 not only because it represents the main receptor for chemokine CCL5 (RANTES) released by platelets following activation, but also because it is an expression of the neutrophil activation state during inflammation and is involved in tissue migration of these cell type [8,9]”.

I am not convinced by the authors that it is appropriate to measure CCR5 on neutrophils. Furthermore, reference 9, which they cite, states that circulating neutrophils have little or no exposure of these receptors, as well as other receptors for chemokines of the CC family. If we are talking about chemokines released from platelets that affect neutrophils, it is primarily IL-8, not RANTES.

The authors made a refinement to the methods. “Flow cytometry data were collected in EDTA-anticoagulated blood samples, within 1 hour after collection, as previously reported”. In my opinion, this is not the most successful option as EDTA, which blocks the activity of integrins, may contribute to the breakdown of leukocyte-platelet aggregates. In my experience and that of other researchers, it is best to take fresh citrate blood and immediately perform a short and gentle prefixing with paraformaldehyde so that the complexes do not fall apart during sample preparation.  Are the authors sure that the complexes did not break down during the one-hour storage of the blood samples?

The authors write “The found percentages of neutrophil positivity (mean ± SD) for CD11b, CD18, CCR5 and CD41a were 98.37 ± 2.27, 64.84 ± 16.95, 8.29 ± 4.42 and 6.51 ± 2.63, respectively. The corresponding RFI values (a.u.) are listed in Table 1….”

I cannot understand why there are significantly fewer CD18+ neutrophils than CD11b+ neutrophils, because the former is not exposed on the cell membrane without CD18.  In general, all neutrophils should have CD11b and CD18. As for neutrophil-platelet aggregates, only some cells form them (~10%). In this case the corresponding dot plots (not histograms) should be provided.  When estimating the relative number of platelets aggregated with neutrophils, fluorescence intensity should be given only for cells stained with antibodies to platelet molecules and not for all neutrophils.

Additionally. Judging from Figure 1, the per cent of CD41+ neutrophils is higher than 6.  

I advise the authors to read the articles discussing the method and results of detection of leukocyte-platelet complexes in human blood samples. doi: 10.1172/jci.insight.153993 https://doi.org/10.1161/hc3801.095588, DOI https://doi.org/ 10.1055/s-0038-1673342. https://doi.org/10.1161/01.CIR.0000015700.27754.6F

Patients with CAD usually take aspirin, some also take clopidogrel, which can influence the formation of neutrophil-platelet aggregates.  The data on therapy could be summarized in a supplementary table. 

Author Response

Comment 1: The authors write “The found percentages of neutrophil positivity (mean ± SD) for CD11b, CD18, CCR5 and CD41a were 98.37 ± 2.27, 64.84 ± 16.95, 8.29 ± 4.42 and 6.51 ± 2.63, respectively. The corresponding RFI values (a.u.) are listed in Table 1….”

I cannot understand why there are significantly fewer CD18+ neutrophils than CD11b+ neutrophils, because the former is not exposed on the cell membrane without CD18. In general, all neutrophils should have CD11b and CD18. As for neutrophil-platelet aggregates, only some cells form them (~10%). In this case the corresponding dot plots (not histograms) should be provided. When estimating the relative number of platelets aggregated with neutrophils, fluorescence intensity should be given only for cells stained with antibodies to platelet molecules and not for all neutrophils.

Response 1: We agree with the Reviewer in highlighting the need for concordance of the percentage of neutrophil expression for CD11b and CD18. Probably, the different fluorescence emission intensity of the corresponding anti-CD11b and anti-CD18 monoclonal antibodies we used, as revealed by the different observed RFI values, affects also the quantification of their percentage of positivity.

Furthermore, as highlighted in Figure 1, in our study there is a strong overlap between the histogram of CD41a+ neutrophils and that of the corresponding control (based uniquely on morphological SSC characteristics vs. CD14low fluorescence intensity). This makes extremely difficult to identify and quantify by a dot-plot the CD41a+ neutrophil cluster. The use of more specific neutrophil identification markers (CD66b and CD16) is, in fact, reported in the limitations of the study.  

---------

Comment 2: I am not convinced by the authors that it is appropriate to measure CCR5 on neutrophils. Furthermore, reference 9, which they cite, states that circulating neutrophils have little or no exposure of these receptors, as well as other receptors for chemokines of the CC family. If we are talking about chemokines released from platelets that affect neutrophils, it is primarily IL-8, not RANTES.

Response 2: Platelets release many chemotactic factors for neutrophils following their activation and binding (to leukocytes and/or to endothelium). We chose to evaluate the expression of CCR5 (RANTES receptor) as an example of a neutrophil surface receptor whose expression is clearly modulated following platelet binding. The need for extension to the evaluation of platelet-modulated expression of chemokine receptors (CXCR1 and CXCR2) involved in the neutrophil functional response to the more specific activating chemokine IL-8, has been highlighted in the limitations of the study.

-----------

Comment 3: The authors made a refinement to the methods. “Flow cytometry data were collected in EDTA-anticoagulated blood samples, within 1 hour after collection, as previously reported”. In my opinion, this is not the most successful option as EDTA, which blocks the activity of integrins, may contribute to the breakdown of leukocyte-platelet aggregates. In my experience and that of other researchers, it is best to take fresh citrate blood and immediately perform a short and gentle prefixing with paraformaldehyde so that the complexes do not fall apart during sample preparation. Are the authors sure that the complexes did not break down during the one-hour storage of the blood samples?

Response 3: The choice of EDTA as an anticoagulant was based on the need to use a single blood sample for a simultaneous monocyte phenotypic evaluation. In particular, the work published by Shantsila E. et al. in the Journal of Thrombosis and Haemostasis 2011, 9, 1056-1066, in which the quantification of percentage of circulating monocyte-platelet aggregates (MPAs) in EDTA-anticoagulated blood was also performed, has been used as a methodological reference [Reference n. 23 cited within the bibliographic reference list (n. 10) of the present submitted work]. However, we agree with the Reviewer regarding the possibility of using a separate gently pre-fixed blood sample anticoagulated with sodium-citrate in order to derive a more precise quantification of neutrophil-platelet aggregates, as well as to avoid a possible dissociation of the heterotypic aggregates during the one-hour storage.

-----------

Comment 4: Additionally. Judging from Figure 1, the per cent of CD41+ neutrophils is higher than 6.

I advise the authors to read the articles discussing the method and results of detection of leukocyte-platelet complexes in human blood samples.

doi: 10.1172/jci.insight.153993

https://doi.org/10.1161/hc3801.095588, DOI https://doi.org/ 10.1055/s-0038-1673342.

https://doi.org/10.1161/01.CIR.0000015700.27754.6

Response 4: The representative example of neutrophil positivity for CD41a shown in Figure 1 refers, in fact, to the highest value found (23.07%). Its inclusion within the series of the other measured values generates the reported mean value of 6.51 ± 2.63 (mean ± SD).  

---------

Comment 5: Patients with CAD usually take aspirin, some also take clopidogrel, which can influence the formation of neutrophil-platelet aggregates. The data on therapy could be summarized in a supplementary table.

Response 5: In the Patients section (2.1) we have indicated the administration of a dose of aspirin (100 mg per day) for the entire selected group.