Sphingosine 1-phosphate receptor-1 signaling enhances neutrophil survival while suppressing inflammation and bacterial host defense functions

Reviewer #1 (Public review):

Summary:

This interesting paper demonstrates that transgenic over-expression of sphingosine 1-phosphate receptor 1 (S1PR1) on neutrophils alters their phenotype, resulting in (1) accumulation of neutrophils in blood, spleen, lung, and liver; (2) a shift in homing receptor expression with reduced CXCR2 and elevated CXCR4; (3) altered transcriptional profile with an increase in "G5c" neutrophils and reduced "module scores" for apoptosis and inflammatory response; (4) reduced ROS production upon fLMP stimulation; and (5) altered responses to bacterial and viral infections of the lung. It raises many interesting questions about how S1P signaling regulates neutrophil biology, and hence will be the basis of future studies. These include: (1) What is the physiological role of S1PR1 signaling in neutrophils? Although there is no dramatic effect on numbers upon S1PR1 loss, is there an effect on any of the other parameters measured? (2) What is unique about the lung that S1PR1 over-expression is particularly impactful there? (3) What distinguishes the bacterial context in which S1PR1 over-expression is maladaptive from the viral context in which S1PR1 over-expression is protective? and (4) Can treatment with an S1PR1 agonist mimic S1PR1 over-expression? As a possibly related question, when in neutrophil development does S1PR1 signaling function to shift the phenotype?

Strengths:

(1) A comprehensive characterization of S1PR1-transgenic neutrophils.

(2) Opens many interesting areas of investigation.

Weaknesses:

Although some characterization of the neutrophil-specific Mrp8-Cre is done, most of the experiments use the more widely expressed LysM-Cre. The redistribution phenotype is much stronger with LysM-Cre than with Mrp8-Cre, so it is unclear what effects are attributable to a cell-intrinsic role of S1PR1, even in studies of neutrophils analyzed ex vivo.

Reviewer #2 (Public review):

The authors have utilised two main models to assess the function of S1PR1 in neutrophils in mice. The knockout of this receptor shows no conclusive effect on neutrophil numbers or functions; it was only the overexpression that resulted in significant alterations. Therefore, often the conclusions do not describe normal or disease physiology but could be useful in a bioengineering context.

Strengths:

From a bioengineering standpoint, this seems like an important study - showing enforced expression of S1PR1 in neutrophils has improved outcomes for influenza infection (Figures 6 and 7).

Weaknesses:

Although the strength is the influenza model, genetic modification of human neutrophils cannot be a strategy, and therefore, is there any way to increase this receptor for mouse, or more importantly, human neutrophils? This study only looks at mice with a non-physiological model of overexpression. It does not offer a real therapeutic option, which drastically hinders the importance of the study. I have other concerns with the data analysis and interpretation, which I detail on a figure-by-figure basis (and how it relates to conclusions) below:

Main specific issues:

(1) Figure 2A+B: This is unconvincing; in the surface staining there seem to be real cells positive for the receptor (high staining in the histogram), but none of the transgenic protein is getting there? This undermines the idea that the effects of the transgene are related to S1P signalling. In the 'Total S1PR1' this is both underwhelming and misleading, as an isotype control (or better S1PR1 knockout) is missing, which would give a better representation of actual expression (flow cytometry autofluorescence famously increases in the red laser channels with fix/perm). The Imagestream chosen images are showing best-case scenarios - and aren't representative. What does the isotype/ KO look like here? All in all, the conclusion on receptor internalization is not well supported, especially when theoretically the TG overexpression should overload S1P availability. This also highlights the lack of another control - does overexpression of another random/non-functional protein have the same effect? To play devil's advocate, perhaps overloading of the ubiquitin-proteasome system is responsible?

(2) Figures 2E-H: In the text, the authors should fix the statement 'Additionally, surface CXCR2 was downregulated and CXCR4 upregulated in LysM-S1pr1 TG neutrophils across bone marrow, spleen, and blood (Fig. 2, E and F)' to better reflect that there is no significant difference in the bone marrow regarding CXCR4. Of note, the total MFI from this data would also be informative, another noticeable absence being the gating strategies for much of the data. Also, alter the statement: 'CD62L expression was largely preserved across compartments, with only a modest reduction in bone marrow neutrophils (Fig. 2G)'. A 50% reduction in CD62L is not modest.

(3) Supplemental Figure 3. A common theme: the wrong statistics have been used here, which has led to a false conclusion. Megakaryocyte/erythrocyte progenitors (MEPs) were only elevated in 2/3 TG mice, and the numbers are so small that this is not significant by any measure of the word. This is certainly not statistically significant if the correct test of (log-normalized) two-way ANOVA is performed (with Sidak's post hoc test). Another acceptable test would be Kruskal-Wallis with Dunn's post-test just for MEPs.

(4) Starting at Figure 3, the authors refer to 'S1PR1hi neutrophil accumulation'. Crucially, the authors must here and throughout be explicitly clear in which cells they are referring to, as this can be misleading - particularly as there are real S1PR1-high cells identified in Figure 2A surface staining. It is my understanding that the authors here mean the transgenic artificially high mice - a very large distinction.

(5) Figure 3A: It is difficult to interpret the figure with the necessary details about the experiment. For instance, there is no mention that this is sterile inflammation or what caused it.

(6) Figure 3B and C: It should be made clear whether these splenic neutrophils are related to the time course of peritoneal inflammation in 3A. Why are there so many apoptotic neutrophils in the spleen? The low numbers here suggest a processing issue rather than real death in vivo (which usually is absent).

(7) Figure 3D: This can also be misleading - the wrong statistics are again used. This should be a log-transformed two-way ANOVA. Regardless of this, the data is not strong enough to be conclusive, a minor effect at best that could also just be related to the type of cell tracker used.

(8) Figure 5E: It is stated that 'LysM-S1pr1 TG mice exhibited a higher bacterial burden in the lungs than controls (Fig. 5E).' Again, misleading results, first the wrong statistical test was used (correct = log norm one-way ANOVA with Tukey's or Kruskal Wallis with Dunn's), secondly the only significance is between S1PR1(fsf) and the Mrp8-S1PR1, not with the LysM TG. 5F is also not strong, with only 2/7 values appearing outside the range of the control - P values can be misleading when poor statistics are used.

(9) Figure 6G: Some discussion should be given for why Neutrophils are lower in BALF in the IAV model - even though higher in the lung in the non-IAC mice in Figure 1. In general, rather than focusing on the non-physiological differences, the discussion could better reflect the inconsistencies and more fully address the difference between the TG and KO and what this means going forward.

Reviewer #3 (Public review):

Summary:

Using mice that overexpress S1PR1 in myeloid cells or specifically in neutrophils, the authors show that increased S1PR1 promotes neutrophil release from the bone marrow and accumulation in blood and peripheral tissues without causing baseline tissue injury. These cells acquire a CXCR4-high, CXCR2-low, CD101-low phenotype, survive longer, and display enhanced mitochondrial metabolism and mTOR signaling, together with reduced apoptotic, inflammatory, and ROS-related programs. Although phagocytosis is preserved, ROS production is markedly reduced. This is associated with impaired bacterial clearance in the lung but improved outcomes during influenza infection, including better survival, less weight loss, improved oxygenation, lower viral burden, and reduced lung inflammation. In contrast, myeloid S1PR1 deletion produces little detectable phenotype. The authors therefore propose that S1PR1 separates neutrophil persistence from inflammatory function, improving tolerance to viral lung injury at the expense of antibacterial defense.

Strengths:

This is a technically solid paper using novel mouse models to overexpress S1PR1 specifically in myeloid cells as well as neutrophils. The data are striking with respect to neutrophil expansion. The diverse roles of neutrophils and their population heterogeneity are an important scientific area that has led to many recent breakthroughs - PMC11785525; PMC12823425, thus this is a timely study.

Weaknesses:

The study mainly demonstrates what S1PR1 overexpression is sufficient to do, rather than establishing the physiological role of endogenous S1PR1. The conclusions should therefore be narrowed unless the authors provide stronger loss-of-function and physiological validation. As written, the abstract ("S1PR1 promotes mitochondrial fitness, enhances survival, and reduces inflammatory output") and the conclusion ("S1PR1 serves as a key regulatory axis") are sufficiency claims but should not be promoted as necessity claims. The honest sentence is: "Thus, a better conclusion would be that enforced S1PR1 expression is sufficient to reprogram neutrophils".

The authors do not confirm efficient S1pr1 deletion in neutrophils. Furthermore, the knockout is examined only under steady-state conditions and limited in vitro stimulation, but not in the bacterial or influenza models where the transgenic phenotype is observed. Without these experiments, the study cannot establish whether endogenous S1PR1 is necessary for the reported functions.

The degree of S1PR1 overexpression is not quantified relative to normal physiological levels. The authors should determine whether naturally occurring S1PR1-high neutrophils display the same survival, metabolic, trafficking, and inflammatory features observed in the transgenic cells.

Analysis of relevant human or mouse datasets, including sepsis, ARDS, viral infection, cancer, or aging, would also help establish whether this neutrophil state exists physiologically.

Surface S1PR1 expression appears similar between control and transgenic neutrophils, whereas total intracellular receptor is increased. This suggests that the phenotype may depend on receptor internalization or endosomal signaling. An internalization-deficient S1PR1 model, such as S1P1-S5A, would help distinguish sustained surface signaling from internalization-dependent signaling. The authors should also determine whether the phenotype requires ligand binding, Gi signaling, and mTOR activity.

The reduction in CXCR2 and decreased neutrophil accumulation in the airways could alone explain the protection from influenza-induced lung injury. The current experiments do not clearly distinguish neutrophil reprogramming from defective migration into the alveolar space.

Although this may be outside the scope of the current study, the authors should directly test whether CXCR2 inhibition reproduces the phenotype.

The reported reduction in viral load should also be confirmed using plaque assay or TCID50, and the possible contribution of NET formation should be examined.

Author response:

Reviewer #1 (Public review):

Summary:

This interesting paper demonstrates that transgenic over-expression of sphingosine 1-phosphate receptor 1 (S1PR1) on neutrophils alters their phenotype, resulting in (1) accumulation of neutrophils in blood, spleen, lung, and liver; (2) a shift in homing receptor expression with reduced CXCR2 and elevated CXCR4; (3) altered transcriptional profile with an increase in "G5c" neutrophils and reduced "module scores" for apoptosis and inflammatory response; (4) reduced ROS production upon fLMP stimulation; and (5) altered responses to bacterial and viral infections of the lung. It raises many interesting questions about how S1P signaling regulates neutrophil biology, and hence will be the basis of future studies. These include: (1) What is the physiological role of S1PR1 signaling in neutrophils? Although there is no dramatic effect on numbers upon S1PR1 loss, is there an effect on any of the other parameters measured? (2) What is unique about the lung that S1PR1 over-expression is particularly impactful there? (3) What distinguishes the bacterial context in which S1PR1 over-expression is maladaptive from the viral context in which S1PR1 over-expression is protective? and (4) Can treatment with an S1PR1 agonist mimic S1PR1 over-expression? As a possibly related question, when in neutrophil development does S1PR1 signaling function to shift the phenotype?

Strengths:

(1) A comprehensive characterization of S1PR1-transgenic neutrophils.

(2) Opens many interesting areas of investigation.

We thank the reviewer for a very positive assessment of our work and for raising very interesting questions, which will be useful to extend this work in the future.

Weaknesses:

Although some characterization of the neutrophil-specific Mrp8-Cre is done, most of the experiments use the more widely expressed LysM-Cre. The redistribution phenotype is much stronger with LysM-Cre than with Mrp8-Cre, so it is unclear what effects are attributable to a cell-intrinsic role of S1PR1, even in studies of neutrophils analyzed ex vivo.

We acknowledge that the data from Mrp8-Cre mouse strain are more limited than the LysM-Cre counterparts. This is because we initially characterized the LysM-Cre S1pr1 KO and TG strains and confirmed key findings relevant to neutrophils in the Mrp8-Cre counterparts. Going forward, more studies will be done in the Mrp8-Cre strain as suggested by the reviewer.

Reviewer #2 (Public review):

The authors have utilised two main models to assess the function of S1PR1 in neutrophils in mice. The knockout of this receptor shows no conclusive effect on neutrophil numbers or functions; it was only the overexpression that resulted in significant alterations. Therefore, often the conclusions do not describe normal or disease physiology but could be useful in a bioengineering context.

We agree with the reviewer that some of the key findings described in our manuscript, for example, neutrophil survival, spleen size, and ROS reduction, etc., were not observed in the S1pr1 KO strains. Our interpretation is that other receptors, for example S1PR4, could be involved in compensating for the loss of S1PR1. We will explain this better in the revisions.

Strengths:

From a bioengineering standpoint, this seems like an important study - showing enforced expression of S1PR1 in neutrophils has improved outcomes for influenza infection (Figures 6 and 7).

We agree with the reviewer that overexpression of S1PR1 could be useful from “bioengineering standpoint”.

Weaknesses:

Although the strength is the influenza model, genetic modification of human neutrophils cannot be a strategy, and therefore, is there any way to increase this receptor for mouse, or more importantly, human neutrophils? This study only looks at mice with a non-physiological model of overexpression. It does not offer a real therapeutic option, which drastically hinders the importance of the study. I have other concerns with the data analysis and interpretation, which I detail on a figure-by-figure basis (and how it relates to conclusions) below:

The reviewer's comments are acknowledged. However, at this early stage of discovery, we feel that it would be premature to address the issue of “a real therapeutic option”. This can be addressed in the future.

Main specific issues:

(1) Figure 2A+B: This is unconvincing; in the surface staining there seem to be real cells positive for the receptor (high staining in the histogram), but none of the transgenic protein is getting there? This undermines the idea that the effects of the transgene are related to S1P signalling. In the 'Total S1PR1' this is both underwhelming and misleading, as an isotype control (or better S1PR1 knockout) is missing, which would give a better representation of actual expression (flow cytometry autofluorescence famously increases in the red laser channels with fix/perm). The Imagestream chosen images are showing best-case scenarios - and aren't representative. What does the isotype/ KO look like here? All in all, the conclusion on receptor internalization is not well supported, especially when theoretically the TG overexpression should overload S1P availability. This also highlights the lack of another control - does overexpression of another random/non-functional protein have the same effect? To play devil's advocate, perhaps overloading of the ubiquitin-proteasome system is responsible?

We will conduct S1PR1 antibody staining with knockout neutrophils as requested by the reviewer. The results will be shown in the revision. The comment of the reviewer on “ubiquitin-proteosome” system is not relevant in our opinion and does not impact the validity of our findings. The approach that we used is standard mouse genetics, and the results should be interpreted from that perspective and not from “devil's advocate”.

(2) Figures 2E-H: In the text, the authors should fix the statement 'Additionally, surface CXCR2 was downregulated and CXCR4 upregulated in LysM-S1pr1 TG neutrophils across bone marrow, spleen, and blood (Fig. 2, E and F)' to better reflect that there is no significant difference in the bone marrow regarding CXCR4. Of note, the total MFI from this data would also be informative, another noticeable absence being the gating strategies for much of the data. Also, alter the statement: 'CD62L expression was largely preserved across compartments, with only a modest reduction in bone marrow neutrophils (Fig. 2G)'. A 50% reduction in CD62L is not modest.

These statements will be changed in the revised manuscript, which we hope to submit soon.

(3) Supplemental Figure 3. A common theme: the wrong statistics have been used here, which has led to a false conclusion. Megakaryocyte/erythrocyte progenitors (MEPs) were only elevated in 2/3 TG mice, and the numbers are so small that this is not significant by any measure of the word. This is certainly not statistically significant if the correct test of (log-normalized) two-way ANOVA is performed (with Sidak's post hoc test). Another acceptable test would be Kruskal-Wallis with Dunn's post-test just for MEPs.

We will reanalyze these data with different statistical tests and discuss these in the revision.

(4) Starting at Figure 3, the authors refer to 'S1PR1hi neutrophil accumulation'. Crucially, the authors must here and throughout be explicitly clear in which cells they are referring to, as this can be misleading - particularly as there are real S1PR1-high cells identified in Figure 2A surface staining. It is my understanding that the authors here mean the transgenic artificially high mice - a very large distinction.

We will clarify and edit this terminology throughout the manuscript in the revision. S1PR1^hi does not strictly imply receptor on the cell surface. It also includes internalized S1PR1 receptors.

(5) Figure 3A: It is difficult to interpret the figure with the necessary details about the experiment. For instance, there is no mention that this is sterile inflammation or what caused it.

This will be edited in the revision.

(6) Figure 3B and C: It should be made clear whether these splenic neutrophils are related to the time course of peritoneal inflammation in 3A. Why are there so many apoptotic neutrophils in the spleen? The low numbers here suggest a processing issue rather than real death in vivo (which usually is absent).

This will be addressed in the revision.

(7) Figure 3D: This can also be misleading - the wrong statistics are again used. This should be a log-transformed two-way ANOVA. Regardless of this, the data is not strong enough to be conclusive, a minor effect at best that could also just be related to the type of cell tracker used.

Same as #3 above.

(8) Figure 5E: It is stated that 'LysM-S1pr1 TG mice exhibited a higher bacterial burden in the lungs than controls (Fig. 5E).' Again, misleading results, first the wrong statistical test was used (correct = log norm one-way ANOVA with Tukey's or Kruskal Wallis with Dunn's), secondly the only significance is between S1PR1(fsf) and the Mrp8-S1PR1, not with the LysM TG. 5F is also not strong, with only 2/7 values appearing outside the range of the control - P values can be misleading when poor statistics are used.

Same as #3 above.

(9) Figure 6G: Some discussion should be given for why Neutrophils are lower in BALF in the IAV model - even though higher in the lung in the non-IAC mice in Figure 1. In general, rather than focusing on the non-physiological differences, the discussion could better reflect the inconsistencies and more fully address the difference between the TG and KO and what this means going forward.

Same as #5 above.

Reviewer #3 (Public review):

Summary:

Using mice that overexpress S1PR1 in myeloid cells or specifically in neutrophils, the authors show that increased S1PR1 promotes neutrophil release from the bone marrow and accumulation in blood and peripheral tissues without causing baseline tissue injury. These cells acquire a CXCR4-high, CXCR2-low, CD101-low phenotype, survive longer, and display enhanced mitochondrial metabolism and mTOR signaling, together with reduced apoptotic, inflammatory, and ROS-related programs. Although phagocytosis is preserved, ROS production is markedly reduced. This is associated with impaired bacterial clearance in the lung but improved outcomes during influenza infection, including better survival, less weight loss, improved oxygenation, lower viral burden, and reduced lung inflammation. In contrast, myeloid S1PR1 deletion produces little detectable phenotype. The authors therefore propose that S1PR1 separates neutrophil persistence from inflammatory function, improving tolerance to viral lung injury at the expense of antibacterial defense.

Strengths:

This is a technically solid paper using novel mouse models to overexpress S1PR1 specifically in myeloid cells as well as neutrophils. The data are striking with respect to neutrophil expansion. The diverse roles of neutrophils and their population heterogeneity are an important scientific area that has led to many recent breakthroughs - PMC11785525; PMC12823425, thus this is a timely study.

We thank the reviewer for positive comments.

Weaknesses:

The study mainly demonstrates what S1PR1 overexpression is sufficient to do, rather than establishing the physiological role of endogenous S1PR1. The conclusions should therefore be narrowed unless the authors provide stronger loss-of-function and physiological validation. As written, the abstract ("S1PR1 promotes mitochondrial fitness, enhances survival, and reduces inflammatory output") and the conclusion ("S1PR1 serves as a key regulatory axis") are sufficiency claims but should not be promoted as necessity claims. The honest sentence is: "Thus, a better conclusion would be that enforced S1PR1 expression is sufficient to reprogram neutrophils".

We will edit the text to better reflect sufficiency versus necessity terms.

The authors do not confirm efficient S1pr1 deletion in neutrophils. Furthermore, the knockout is examined only under steady-state conditions and limited in vitro stimulation, but not in the bacterial or influenza models where the transgenic phenotype is observed. Without these experiments, the study cannot establish whether endogenous S1PR1 is necessary for the reported functions.

We will provide qRT-PCR data (that we have done already but did not include in the original version) confirming efficient deletion of the S1pr1 gene.

The degree of S1PR1 overexpression is not quantified relative to normal physiological levels. The authors should determine whether naturally occurring S1PR1-high neutrophils display the same survival, metabolic, trafficking, and inflammatory features observed in the transgenic cells.

We agree that the level of S1PR1 overexpression in our transgenic neutrophils should be quantitatively compared with endogenous S1PR1 expression. We will provide the quantitative result in the revision. Our analyses of independent human (GSE216009) and mouse (GSE243466 and GSE266518) single-cell RNA-seq datasets indicate that endogenous mRNA expression for this receptor is higher in specific neutrophil states associated with inflammatory, immature, and tissue-associated populations. Notably, these endogenous S1PR1-expressing neutrophils show transcriptional features involving altered oxidative/inflammatory programs, chemotaxis, and mitochondrial/metabolic regulation that partially overlap with the phenotype of S1PR1-transgenic neutrophils. We will provide additional quantitative and dataset analyses in the revised manuscript.

Analysis of relevant human or mouse datasets, including sepsis, ARDS, viral infection, cancer, or aging, would also help establish whether this neutrophil state exists physiologically.

As described above, we have analyzed independent human and mouse single-cell RNA-seq datasets from sepsis, cancer, and other inflammatory conditions to determine whether endogenous S1PR1-expressing neutrophil states occur naturally across biological and disease contexts. These analyses support the presence of distinct endogenous S1PR1-expressing neutrophil populations across multiple biological contexts. We will provide the expanded analyses and corresponding data in the revised manuscript.

Surface S1PR1 expression appears similar between control and transgenic neutrophils, whereas total intracellular receptor is increased. This suggests that the phenotype may depend on receptor internalization or endosomal signaling. An internalization-deficient S1PR1 model, such as S1P1-S5A, would help distinguish sustained surface signaling from internalization-dependent signaling. The authors should also determine whether the phenotype requires ligand binding, Gi signaling, and mTOR activity.

We agree that future experiments will use internalization-defective S1PR1 S5A knock-in neutrophils.

The reduction in CXCR2 and decreased neutrophil accumulation in the airways could alone explain the protection from influenza-induced lung injury. The current experiments do not clearly distinguish neutrophil reprogramming from defective migration into the alveolar space.

We agree. We did not examine the role of CXCR2 in the influenza experiments. Our interpretation is that reduced ROS from transgenic neutrophils reduced lung injury.

Although this may be outside the scope of the current study, the authors should directly test whether CXCR2 inhibition reproduces the phenotype.

This is a good point that the reviewer brought up. This can be addressed in a future study.

The reported reduction in viral load should also be confirmed using plaque assay or TCID50, and the possible contribution of NET formation should be examined.

This is again a good suggestion that can be addressed in a future study.