A neutrophil–B-cell axis impacts tissue damage control in a mouse model of intraabdominal bacterial infection via Cxcr4

  1. Riem Gawish
  2. Barbara Maier
  3. Georg Obermayer
  4. Martin L Watzenboeck
  5. Anna-Dorothea Gorki
  6. Federica Quattrone
  7. Asma Farhat
  8. Karin Lakovits
  9. Anastasiya Hladik
  10. Ana Korosec
  11. Arman Alimohammadi
  12. Ildiko Mesteri
  13. Felicitas Oberndorfer
  14. Fiona Oakley
  15. John Brain
  16. Louis Boon
  17. Irene Lang
  18. Christoph J Binder
  19. Sylvia Knapp

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This experiment investigated the link between the role of B lymphocytes to neutrophils for the achievement of LPS tolerance. The authors found that B cells can modulate the tissue-damaging properties of neutrophil leukocytes by influencing neutrophil Cxcr4 signaling in a mouse model of bacterial sepsis.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Sepsis is a life-threatening condition characterized by uncontrolled systemic inflammation and coagulation, leading to multiorgan failure. Therapeutic options to prevent sepsis-associated immunopathology remain scarce. Here, we established a mouse model of long-lasting disease tolerance during severe sepsis, manifested by diminished immunothrombosis and organ damage in spite of a high pathogen burden. We found that both neutrophils and B cells emerged as key regulators of tissue integrity. Enduring changes in the transcriptional profile of neutrophils include upregulated Cxcr4 expression in protected, tolerant hosts. Neutrophil Cxcr4 upregulation required the presence of B cells, suggesting that B cells promoted disease tolerance by improving tissue damage control via the suppression of neutrophils’ tissue-damaging properties. Finally, therapeutic administration of a Cxcr4 agonist successfully promoted tissue damage control and prevented liver damage during sepsis. Our findings highlight the importance of a critical B-cell/neutrophil interaction during sepsis and establish neutrophil Cxcr4 activation as a potential means to promote disease tolerance during sepsis.

Article activity feed

  1. Version published to 10.7554/elife.78291 on eLife
  2. eLife

    Evaluation Summary:

    This experiment investigated the link between the role of B lymphocytes to neutrophils for the achievement of LPS tolerance. The authors found that B cells can modulate the tissue-damaging properties of neutrophil leukocytes by influencing neutrophil Cxcr4 signaling in a mouse model of bacterial sepsis.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  3. eLife

    Reviewer #1 (Public Review):

    This is a study linking the role of B lymphocytes to neutrophils for the achievement of LPS tolerance in an experimental setting. The manuscript is elegantly written and easy to follow. One main strength of this submission is the extensive mechanistic insights involving even transfusion of splenocytes.

  4. eLife

    Reviewer #2 (Public Review):

    The topic is interesting, the study addresses an important question, and the manuscript is fairly well written. However, in my opinion, there are a number of serious problems that need to be addressed - because there are many similar laboratory studies on this subject. In my opinion, the novelty of the main message (a neutrophil - B-cell axis governs disease tolerance during sepsis via Cxcr4) is limited. My explanation is as follows.

    Firstly, human sepsis is defined (according to Sepsis 3 criteria) as life-threatening organ dysfunction caused by a dysregulated host response to infection (Singer et al. 2016) and in agreement with the Authors, despite considerable research efforts the pathophysiology is still unknown. Besides, the mechanisms underlying the signaling and trafficking of PMN leukocytes within the affected tissues, and the roles of adhesion molecules, including chemokine receptors are still not fully understood. Indeed, diverse therapies directed against these targets have shown dramatic effects in animal models; however, in humans, their clinical impact has been modest. There are many reasons for the discordance between biological promise and bedside reality, and one of them might be the use of animal models. Extrapolation from animal models always requires an awareness of species-related and other dependent variations and I would like to stress that this study employed mice, the Author has investigated the mechanisms involved in the PMN reactions in a mouse model of intraabdominal (O18:K1 E. Coli) bacterial infection, but this fact is not mentioned in the title or abstract (only in keywords) or introduction.

    Secondly, according to the Authors they "established and investigated a model of disease tolerance during sepsis, which enabled them to reveal the importance of B cells and neutrophils in mediating tissue tolerance in the context of severe infections". Here it should be noted that 1. LPS administration to rodents does not mimic human sepsis, and 2. the mechanism of endotoxin tolerance (or preconditioning) has been studied for decades. It is characterized by a hyporesponsive state following low-dose stimulation with a TLR4 ligand (i.e. lipopolysaccharide). Besides, endotoxin tolerance provokes cross-tolerance against other forms of injuries as well such as liver ischemia-reperfusion (J Surg Res 57, 1994).

    Thirdly, according to the Authors they considered the possibility that B cells regulate infection-induced neutrophil functionalities because "we discovered that LPS-induced protection was still observed in splenectomized animals". Nevertheless, the role of spleen in endotoxin effects (or more correctly, that the absence of spleen) in the development of tolerance to endotoxin was demonstrated decades ago (and studied repeatedly by the group of Agarwal (Br J Exp Pathol 53 1972).

    Next, according to the Authors, "it is tempting to speculate that B cells act as important regulators of granulopoiesis and neutrophil trafficking at steady state and under inflammatory conditions". Nevertheless, previous studies have established that neutrophil trafficking is regulated mainly via CXCR4 in steady state, and the attenuation of CXCR4 signaling leads to the entry of PMNs into the circulation from the bone marrow (JCI 120(7) 2010).

    Furthermore, selective inhibition of CXCR4 by AMD3100 is protective in many injury models, including mice with fecal peritonitis (Front. Immunol 2020 | https://doi.org/10.3389/fimmu.2020.00407).

    Lastly, it has been previously recorded that Rag1 animals had a higher death rate to CLP than wild type animals and that this could be ameliorated by adaptive transfer of syngeneic T cells induced to overexpress the anti-apoptotic molecule Bcl-2 by gene transfer techniques (Hotchkiss, Crit Care Med 1997)

    Therefore, until this point, I consider the work largely repetitive and suggest to downgrade the wording a bit due to the lack of novelty. The paper confirms these previous observations in a partially new context, which is the demonstration of a crosstalk between PMNs and B cells in the bone marrow, in which B cells influence PMN trafficking likely by modulating Cxcr4 related pathways.

  5. eLife

    Author Response:

    Reviewer #2 (Public Review):

    The topic is interesting, the study addresses an important question, and the manuscript is fairly well written. However, in my opinion, there are a number of serious problems that need to be addressed - because there are many similar laboratory studies on this subject. In my opinion, the novelty of the main message (a neutrophil - B-cell axis governs disease tolerance during sepsis via Cxcr4) is limited. My explanation is as follows.

    Firstly, human sepsis is defined (according to Sepsis 3 criteria) as life-threatening organ dysfunction caused by a dysregulated host response to infection (Singer et al. 2016) and in agreement with the Authors, despite considerable research efforts the pathophysiology is still unknown. Besides, the mechanisms underlying the signaling and trafficking of PMN leukocytes within the affected tissues, and the roles of adhesion molecules, including chemokine receptors are still not fully understood. Indeed, diverse therapies directed against these targets have shown dramatic effects in animal models; however, in humans, their clinical impact has been modest. There are many reasons for the discordance between biological promise and bedside reality, and one of them might be the use of animal models. Extrapolation from animal models always requires an awareness of species-related and other dependent variations and I would like to stress that this study employed mice, the Author has investigated the mechanisms involved in the PMN reactions in a mouse model of intraabdominal (O18:K1 E. Coli) bacterial infection, but this fact is not mentioned in the title or abstract (only in keywords) or introduction.

    We thank the reviewer for this comment. We fully agree that the translational potential of many mouse studies is limited, but truly believe that there are conserved mechanisms between mouse and human which are worth being explored and reported. We certainly did not intend to hide that our study was performed in mice and have now clarified this by adapting the abstract as follows: “Here, we established a mouse model of long-lasting disease tolerance during severe sepsis, manifested by diminished immunothrombosis and organ damage in spite of a high pathogen burden.“ page 2, line 11

    Secondly, according to the Authors they "established and investigated a model of disease tolerance during sepsis, which enabled them to reveal the importance of B cells and neutrophils in mediating tissue tolerance in the context of severe infections". Here it should be noted that 1. LPS administration to rodents does not mimic human sepsis, and 2. the mechanism of endotoxin tolerance (or preconditioning) has been studied for decades. It is characterized by a hyporesponsive state following low-dose stimulation with a TLR4 ligand (i.e. lipopolysaccharide). Besides, endotoxin tolerance provokes cross-tolerance against other forms of injuries as well such as liver ischemia-reperfusion (J Surg Res 57, 1994).

    Ad 1. We thank the reviewer for this critical comment. While we fully agree on the fact that LPS administration does not mimic human sepsis, we are slightly puzzled by this criticism as we at no point say so in our manuscript. What we instead say is that E. coli peritonitis is a sepsis model, a claim that is widely accepted in the field.

    We are explaining this in the intro: “In this study, we investigated mechanisms of disease tolerance and tissue damage control, by comparing tolerant and sensitive hosts during a severe bacterial infection. While sensitive animals developed severe coagulopathy and tissue damage during sepsis, tolerant animals were able to maintain tissue integrity in spite of a high bacterial load. Disease tolerance was induced by the prior exposure of animals to a single, low-dose of LPS and could be uncoupled from LPS-induced suppression of cytokine responses.” page 3, line 32-34, page 4, line 1-3

    …as well as in the result section: “We thus challenged mice intravenously (i.v.) with a subclinical dose of LPS 1 day, 2 weeks, 5 weeks or 8 weeks, respectively, prior to the induction of Gram-negative sepsis by intraperitoneal (i.p.) injection of the virulent E. coli strain O18:K1.” page 5, line 7-9

    … and in our methods section, which we have adapted to make this point even more clear to the reader: Tolerance was induced by i.v. injection of 30μg E. coli LPS (Sigma-Aldrich) at indicated times before induction of bacterial sepsis by intraperitoneal (i.p.) infection with 1-2x104 E. coli O18:K1. E. coli peritonitis was induced as described previously (Knapp, de Vos et al. 2003, Knapp, Matt et al. 2007, Gawish, Martins et al. 2015). page 24, line 13-15

    Ad 2. We agree that LPS- or endotoxin tolerance is an old topic which has been studied for a long time. However, as we explain in detail in our response to the editor, the contribution of “LPS tolerance” to “disease tolerance” is still under investigation. As explained extensively above (please refer to our response to the editor), we observe signs of LPS tolerance in our experimental setup, as LPS pre-exposed mice produce lower levels of inflammatory cytokines shortly after infection. However, our data collectively suggest, that reduced early cytokine production (most likely a sign of LPS tolerance) is not the reason for improved tissue damage control in LPS-pre-treated mice. As such, we argue that the protective phenotype we observed is independent of early cytokine suppression and independent of monocytes and macrophages, hence not a result of LPS tolerance.

    Thirdly, according to the Authors they considered the possibility that B cells regulate infection-induced neutrophil functionalities because “we discovered that LPS-induced protection was still observed in splenectomized animals”. Nevertheless, the role of spleen in endotoxin effects (or more correctly, that the absence of spleen) in the development of tolerance to endotoxin was demonstrated decades ago (and studied repeatedly by the group of Agarwal (Br J Exp Pathol 53 1972).

    It seems that there has been a misunderstanding as we do not make this claim in our manuscript. To be precise the correct wording that we used in our manuscript is: “Since we discovered that LPS-induced protection was still observed in splenectomized animals, we considered the possibility that B cells regulate infection-induced neutrophil functionalities via effects exerted by sharing the same bone marrow niche. In fact, B cells, neutrophils and their precursors build up the majority of the constitutive CD45+ bone marrow cell pool, where they mature while sharing the same niche (Yang, Busche et al. 2013).“ page 19, line 17-21

    This obviously has a different meaning. As we saw that LPS-induced tissue tolerance was abrogated in (full body) B cell deficient mice, our intention was not to study the effect of splenectomy on LPS-tolerance but to make use of splenectomy to narrow down the B cell compartment which is driving the protective effect. Of course, the role of the spleen during endotoxemia and during peritonitis has been studied before, but the point we wanted to make with our finding was that splenectomy, in contrast to a full B cell deficiency (in Rag2-/- or JHT-/- animals), did not abrogate LPS-induced tissue protection.

    To address and clarify this point, we slightly modified our explanation in the results section and included some of these “old” splenectomy studies: “We then tested if splenectomy would replicate the protective effects of full B cell deficiency during sepsis and interestingly found that splenectomy was associated with reduced liver damage in naïve, sensitive mice, which is in line with other studies (Agarwal, Parant et al. 1972, Karanfilian, Spillert et al. 1983), but, in contrast to complete lymphocyte deficiency, not sufficient to abrogate LPS-induced tissue protection in tolerant animals (Figure 2G and S2F). This suggested that mature splenic B cells contributed to tissue damage during severe infections, while other, not spleen derived, B cell compartments were instrumental in driving disease tolerance.“ page 8, line 1-7

    Next, according to the Authors, "it is tempting to speculate that B cells act as important regulators of granulopoiesis and neutrophil trafficking at steady state and under inflammatory conditions". Nevertheless, previous studies have established that neutrophil trafficking is regulated mainly via CXCR4 in steady state, and the attenuation of CXCR4 signaling leads to the entry of PMNs into the circulation from the bone marrow (JCI 120(7) 2010).

    We believe that there is another misunderstanding. This comment suggests that an involvement of B cells in neutrophil regulation would be against already published knowledge about CXCR4 as a master regulator of neutrophil trafficking. This is not at all what we are claiming in our manuscript.

    A potential involvement of B cells in neutrophil regulation is not in conflict with what is already known about the important role of CXCR4. As CXCR4 signaling is critical for both B cells AND neutrophils, competition for the CXCR4 ligand Cxcl12 (SDF1) and differences in the sensitivity to altered ligand availability might serve as an explanation.

    We are actually discussing the role of CXCR4 in neutrophil trafficking extensively in the results section and in the discussion of our manuscript and have now slightly modified our wording and extended our discussion part to clarify this point:

    “…neutrophil aging, a process that is counteracted by Cxcr4 signaling, the master regulator of neutrophil trafficking between the bone marrow and the periphery (Martin, Burdon et al. 2003, Eash, Greenbaum et al. 2010, Adrover, Del Fresno et al. 2019).” page 13, line 7-9

    “Considering the reported importance of Cxcr4 signaling in neutrophil retention in the bone marrow and their release to the periphery (Martin, Burdon et al. 2003, Eash, Greenbaum et al. 2010, Adrover, Del Fresno et al. 2019),…“ page 15, line 1-2

    “Cxcr4 interaction with its ligand Cxcl12 (stromal cell-derived factor 1, SDF1) has been shown to be critical for the retention of neutrophils in the bone marrow under steady state, their release to the periphery as well as their homing back to the bone marrow when they become senescent (Martin, Burdon et al. 2003, Eash, Greenbaum et al. 2010). Importantly, Cxcr4 signaling is essential, as Cxcr4 knockout mice die perinatally due to severe developmental defects ranging from virtually absent myelopoiesis and impaired B lymphopoiesis to abnormal brain development (Ma, Jones et al. 1998). A different sensitivity to changes in SDF1 concentrations as a potential mechanism of the reciprocal regulation of lymphopoiesis and granulopoiesis has been suggested earlier (Ueda, Kondo et al. 2005).“ page 19, line 33-34, page 20, line 1-7

    Furthermore, selective inhibition of CXCR4 by AMD3100 is protective in many injury models, including mice with fecal peritonitis (Front. Immunol 2020 | https://doi.org/10.3389/fimmu.2020.00407).

    We thank the reviewer for pointing this out. In the above cited paper injection of AMD3100 blocks neutrophil migration into the peritoneal cavity as well as into the tissue during zymosan- or fecal slurry-induced peritonitis. Interestingly, we do not see any benefit of Cxcr4 inhibition (using AMD3100) in our model, but in contrast observed tissue protection by administration of the Cxcr4 agonist ATI2341 (Figure 5F). This can be explained by a different timing (and maybe also route) of administration. In the aforementioned study, AMD3100 is injected intraperitoneally, 1h BEFORE induction of peritonitis. AMD3100 injection is known to cause an immediate release of neutrophils from the bone marrow to the blood (Liu, Li et al. 2015), which means that in this experimental setup, neutrophils are already in the circulation prior the injection of zymosan or fecal slurry. Our experimental setup in contrast uses a therapeutic approach, as we inject AMD3100 or ATI2341 just 6h post induction of E. coli peritonitis. At 6h post infection, E. coli infection has already caused a massive increase in blood neutrophils and infiltration of neutrophils into the peritoneum and various organs. We thus believe that these experimental setups are not truly comparable, but agree with the reviewer that this needs to be addressed in our manuscript.

    We have now added the following paragraph to the discussion to address this study:

    “Given its clinical importance, Cxcr4 inhibition (using AMD3100) has been studied in different injury models, but interestingly only little is known about the therapeutic impact of Cxcr4 activation. Strikingly, activating, but not antagonizing, Cxcr4 during sepsis promoted tissue damage control in our model, which is in conflict with a study showing that Cxcr4 blockade with AMD3100 prior induction of peritonitis prevents neutrophil infiltration and tissue inflammation (Ngamsri, Jans et al. 2020). While we only see a tissue protective effect of ATI2341, but not AMD3100, we believe that this is due to differences in the timing and maybe also the route of drug administration. As we use a therapeutic approach and target Cxcr4 as late as 6h post E. coli injection, a time when there is already substantial neutrophilia in blood and organs, our data support an impact of Cxcr4 signaling on neutrophils’ tissue damaging properties and suggest that B cell driven regulation of Cxcr4 is a potential mechanism of disease tolerance and thus might be an interesting therapeutic target during severe sepsis.“ page 20, line 12-23

    Lastly, it has been previously recorded that Rag1 animals had a higher death rate to CLP than wild type animals and that this could be ameliorated by adaptive transfer of syngeneic T cells induced to overexpress the anti-apoptotic molecule Bcl-2 by gene transfer techniques (Hotchkiss, Crit Care Med 1997)

    We thank the reviewer for this comment. We are aware of the fact that the role of lymphocytes in different experimental models of peritonitis-induced sepsis has been studied before. However, the contribution of different lymphocyte populations remains controversial, likely due to different experimental setups and methodologies used in different studies. While we do not really see the connection of lymphocyte apoptosis to our study, we want to point out that the study, which is mentioned by the reviewer does not report a higher death rate of Rag1-/- animals, but shows that there is substantial lymphocyte apoptosis during sepsis (Hotchkiss, Swanson et al. 1997). We believe that the study which the reviewer refers to is from 1999 and has been published in the Journal of Immunology (Hotchkiss, Swanson et al. 1999).

    However, as lymphocyte apoptosis is not connected to the topic of our study, we chose to discuss our data in the context of more B cell specific literature and have cited another important CLP study which showed that IFN-activated B cells are critical for survival by promoting early inflammation and neutrophil effector functions during CLP which in turn improves bacterial clearance (Kelly-Scumpia, Scumpia et al. 2011). In this setup, the increased mortality of Rag1-/- animals during CLP is likely a result of an increased pathogen load and therefore not in conflict with our data. While we appreciate that CLP is the more physiologic model, the strength of our E. coli sepsis model is that bacterial outgrowth already occurs at a maximum speed and is not further enhanced by the absence of certain immune cell types which is why we can uncouple immunopathologic effects from the pathogen load.

    We mention this study in our result section: “Given that B cells were shown to promote early production of proinflammatory cytokines such as IL-6 during sepsis in a type I IFN dependent manner …“ page 8, line 8-9

    We have further added a paragraph to our discussion to better explain the differences between CLP and the model we have used for this study: “It was demonstrated earlier that mature, splenic B2 cells promote neutrophil activation by boosting type-I IFN dependent early inflammation, which in turn improves bacterial clearance and survival during CLP (Kelly-Scumpia, Scumpia et al. 2011). While enhanced inflammation can mediate pathogen clearance during CLP, it at the same time contributes to tissue damage which is of particular importance in our model. In support of proinflammatory, tissue-damaging properties of mature B2 cell subsets, we found splenectomy similarly protective as B cell deficiency during primary sepsis and reconstitution of Rag2-/- mice with B cells to increase tissue damage. Interestingly, we did not identify an important role for the proposed IFNAR-driven inflammatory function of B cells (Kelly-Scumpia, Scumpia et al. 2011) in sepsis, and inflammation did not differ between wild type and lymphocyte deficient mice. “ page 18, line 29-34, page 19, line 1-5

    Therefore, until this point, I consider the work largely repetitive and suggest to downgrade the wording a bit due to the lack of novelty. The paper confirms these previous observations in a partially new context, which is the demonstration of a crosstalk between PMNs and B cells in the bone marrow, in which B cells influence PMN trafficking likely by modulating Cxcr4 related pathways. We have highlighted the novelties of our study in this response letter and our revised manuscript, and hope the reviewer agrees with us that this improved the clarity and better explains the novelties.

  9. Version published to 10.1101/2022.03.21.485114v1 on bioRxiv