Neutrophils as drivers of vascular injury in sickle cell disease

Summary

While neutrophils are the main effectors of protective innate immune responses, they are also key players in inflammatory pathologies. Sickle cell disease (SCD) is a genetic blood disorder in which red blood cells (RBCs) are constantly destroyed in the circulation which generates a highly inflammatory environment that culminates in vascular occlusions. Vaso-occlusion is the hallmark of SCD and a predictor of disease severity. Neutrophils initiate and propagate SCD-related vaso-occlusion through adhesive interactions with the activated and dysfunctional endothelium, sickle RBCs, and platelets, leading to acute and chronic complications that progress to irreversible organ damage and ultimately death. The use of SCD humanized mouse models, in combination with in vivo imaging techniques, has emerged as a fundamental tool to understand the dynamics of neutrophils under complex inflammatory contexts and their contribution to vascular injury in SCD. In this review, we discuss the various mechanisms by which circulating neutrophils sense and respond to the wide range of stimuli present in the blood of SCD patients and mice. We argue that the central role of neutrophils in SCD can be rationalized to develop targets for the management of clinical complications in SCD patients.

1. INTRODUCTION

Neutrophils are key components of the innate immune system and the first responders to infection and inflammation ¹. Approximately 10¹¹ neutrophils are generated every day in the bone marrow compartment, where they undergo different stages of development from myeloid progenitor cells to mature neutrophils ^2,3. The granulocyte colony stimulating factor (G-CSF) is the main regulator of granulopoiesis and release of mature neutrophils into the bloodstream ^4–6. G-CSF induces the commitment of progenitor cells to the myeloid lineage, promotes proliferation of neutrophil precursors, and reduces the transit time through the bone marrow compartment ^1,7. After their release from the bone marrow, neutrophils circulate for about half-day before migrating to tissues, such as liver, lung and spleen, or homing back to the bone marrow, with consequences in regulating the hematopoietic niches within ⁸.

During infections or upon inflammatory stimuli, granulopoiesis is increased in the bone marrow and a large number of neutrophils migrate to the site of infection or injury. In this process, their lifespan, originally short, is believed to be considerably extended ⁹. Acute inflammation is a highly coordinated process modulated by inflammatory mediators released by sentinel immune cells and the endothelium, followed by a resolution phase that reestablishes homeostasis ¹⁰. However, when the inflammatory stimulus is not removed or is constantly generated, this resolution phase fails and creates a chronic inflammatory state that drives organ pathology. This is the case of atherosclerosis, cancer, sterile injury, diabetes mellitus, autoimmune diseases and, as detailed below, sickle cell disease. In all of these disorders neutrophils have been demonstrated to play a prominent pathogenic role ^11,12.

The endothelium is a physical barrier between blood cells and the tissue parenchyma that is composed by a highly dynamic monolayer of cells located in the inner face of the vessels. These endothelial cells continuously respond to the extracellular environment and play an extensive role in immune responses and maintenance of vascular hemostasis. The endothelium controls vessel wall permeability, regulates vascular pressure and blood flow, establishes the balance of pro- and anticoagulants, and regulates inflammation signaling ¹³. Constant activation of endothelial cells by excessive inflammatory stimuli, however, promotes endothelial damage and a gradual loss of function. While a healthy endothelium has vaso-dilation capacity and anti-thrombotic and anti-inflammatory functions preserved, a dysfunctional endothelium promotes vaso-constriction and presents a pro-inflammatory and pro-thrombotic state, which can irreversibly damage the vasculature and the closely irrigated organs ¹⁴.

Sickle cell disease (SCD), a genetic blood disorder, is a classic example of chronic (non-resolving) inflammatory condition with progressive multiorgan damage, in which neutrophil-endothelium interactions are key determinants of disease severity and mortality ^15–17. Over the past two decades, neutrophils have emerged as protagonists of SCD pathophysiology. The severely altered environment promoted by SCD creates a unique model of inflammatory disease accompanied by vasculopathy in which neutrophils may display a broad spectrum of phenotypes and functions. In this review, we explore neutrophil biology and elucidate some of the fascinating features of these cells under the prism of vascular occlusions in SCD.

2. SICKLE CELL DISEASE AS AN INFLAMMATORY DISEASE

The most common and severe form of SCD is caused by the homozygous inheritance of a point mutation in the beta-globin gene (HBB), affecting the beta-globin chains that form the hemoglobin tetramer ¹⁸. The normal adult hemoglobin (HbA) is formed by a tetramer of two alpha (α₂) and two beta (β₂) globin chains (α₂β₂), while the point mutation that leads to SCD results in the abnormal hemoglobin S (HbS, α₂β^S₂). The pathophysiology of SCD is triggered by the high propensity of HbS to form polymers in red blood cells (RBCs) when deoxygenated, leading to severe cellular alterations in RBC, such as abnormal rheology, dehydration, expression of adhesion molecules, and the acquisition of a stiffed and sickled shape, the major cellular feature of mutation disease¹⁹. Abnormalities in RBCs caused by HbS polymerization manifest in a vicious cycle of hemolysis and vascular occlusions, the hallmarks of this disease. Shortening of sickle RBCs lifespan results in the destruction of 10% of total RBCs per day, leading to chronic hemolytic anemia. Approximately one-third of sickle-related hemolysis occurs intravascularly, with release of cell-free hemoglobin into the blood that directly damages the vasculature, generates ROS, and triggers various inflammatory pathways (Fig. 1A)^20,21.

Figure 1. Current paradigm of vaso-occlusion in sickle cell disease.

(A) Hemolysis leads to the release of damage associated molecular patterns (DAMPs) from red blood cells (RBCs) into the blood. Free hemoglobin (Hb) scavenges endothelial nitric oxide (NO), reducing its availability to smooth muscle cells and impairing blood vessel dilation. Hb oxidation forms methemoglobin (MetHb), a highly reactive molecule that generates reactive oxygen species (ROS) and releases free heme. (B) The endothelial damage is perpetuated by direct activation of endothelial cells by heme and ROS. The activated endothelium overexpresses adhesion molecules and releases inflammatory cytokines to the circulation, which in turn recruit and activate neutrophils. Neutrophils are also activated by heme and ROS. This cascade of reactions leads to inflammation and vaso-occlusion. (C) Neutrophils are the central effectors of vaso-occlusion due to adhesive interactions with RBCs, platelets and the endothelium, mediated by adhesion molecules as illustrated. PSGL-1, P-selectin glycoprotein ligand-1; ESL-1, E-selectin ligand-1; Mac-1, α_Mβ₂ integrin (CD11b/CD18); LFA-1, α_Lβ₂ integrin (CD11b/CD18); ICAM-1, intercellular adhesion molecule-1; VCAM-1, vascular cell adhesion molecule-1; GPIba, glycoprotein Ib alpha.

Extracellular oxygenated hemoglobin (Oxy-Hb-Fe²⁺) depletes endothelial nitric oxide (NO) in a rapid and irreversible reaction that forms methemoglobin (Hb-Fe³⁺) and nitrate (NO³⁻) ^22,23. The compromised bioavailability of endothelial-derived NO to the adjacent smooth muscle cells impairs vascular homeostasis, leading to vaso-constriction and endothelial dysfunction, with subsequent leukocyte activation and adhesion, and increased platelet activity, adhesion and aggregation (Fig. 1B) ²². RBC lysis also releases arginase, which catabolizes arginine, the substrate for NO synthesis ²⁴. Additionally, the oxidized hemoglobin (methemoglobin, Hb-Fe³⁺) releases free heme, the major erythrocyte-derived damage associated molecular pattern (DAMP) driving sterile inflammation in SCD ²⁵. Cell-free heme into plasma generates ROS, propagating oxidative stress and endothelial activation, and stimulates the innate immune system through toll-like receptor 4 (TLR4) and inflammasome signaling ^16,26,27.

The highly inflammatory environment in the vasculature of SCD patients is evidenced by high levels of proinflammatory cytokines in circulation, including tumor necrosis factor alpha (TNFα), interleukin-6 (IL-6), and interleukin-1 beta (IL-1β) ^28,29, which promote and perpetuate endothelial and leukocyte activation. As discussed later, this proinflammatory milieu causes heterocellular adhesive interactions of sickle RBCs and activated leukocytes and platelets with the activated endothelium, which delay and obstruct the blood flow, promoting vascular occlusions (Fig. 1B). Ischemia-reperfusion episodes lead to acute tissue damage, and their recurrence drives progressive and irreversible organ damage, ultimately leading to multiple organ failure and, in some instances, death ^16,30. Acute vaso-occlusive events accompanied by severe pain are the main cause of hospitalization in SCD patients ³¹ and often evolve to more severe complications, such as acute chest syndrome, a severe type of lung injury that represents the leading cause of death among SCD patients ^32,33.

Hydroxyurea, also known as hydroxycarbamide, is the most widely used drug currently available to treat and prevent the acute complications of SCD. Hydroxyurea is a cytostatic drug that strongly induces an increase in fetal hemoglobin (HbF), which consequently dilutes the concentration of HbS into the RBCs, preventing HbS polymerization, RBC sickling, and the subsequent cascade of events that lead to vaso-occlusion ^34,35. HbF is predominantly produced during fetal life, and is progressively replaced by HbA after birth, or by HbS in SCD patients ^36,37. High levels of HbF in adult life of SCD patients ameliorate the clinical course by reducing the frequency of vaso-occlusive crisis, acute chest syndrome, blood transfusions, and hospitalization ^38–40. The beneficial effects of hydroxyurea, however, are not restricted to the induction of HbF, but also coincide with a reduction in leukocyte counts, particularly neutrophils, as supported by the observation that amelioration of clinical symptoms occurs even in the absence of raised HbF ³⁸. Evidence from in vivo and in vitro studies point to hydroxyurea as a potent anti-inflammatory drug that promotes a decrease in neutrophil counts, prevents neutrophil activation and adhesion, and improves endothelial function ^41,42. New drug therapies have been approved by the FDA within the last five years, including L-glutamine ⁴³, an amino acid with anti-oxidant properties; crizanlizumab ⁴⁴, a humanized antibody targeting P-selectin, which expressed by platelets and endothelial cells; and voxelotor ⁴⁵, a modulator of hemoglobin-oxygen affinity. Besides targeting distinct pathways of SCD pathophysiology, they all directly or indirectly reduce leukocyte count, activation and adhesion, further supporting the prominent inflammatory aspect of SCD and the fundamental role of leukocytes to disease pathophysiology.

3. NEUTROPHILS IN SICKLE CELL DISEASE

Elevated leukocyte count is a feature of SCD ^46,47 and a risk factor for disease severity ^48,49. High baseline leukocyte counts have been associated with death at young age ⁵⁰, increased susceptibility to acute chest syndrome ⁵¹, and high frequency of hemorrhagic stroke ⁵². Numerous clinical observations support the relevance of neutrophils to SCD pathophysiology. Evidence from G-CSF treatment used to mobilize hematopoietic stem cells in a SCD patient showed an increase in neutrophil count followed by severe painful vaso-occlusive crisis and acute chest syndrome, with resolution after stopping G-CSF and re-introducing hydroxyurea ⁵³. Subsequent studies also reported acute multiorgan failure ⁵⁴ and even death ⁵⁵ after G-CSF administration to patients with SCD.

In vivo studies using transgenic murine models of SCD expressing human HbS revealed that neutrophils are crucial players in the vaso-occlusive process. Intravital microscopy of the microvasculature of the cremaster muscle in SCD mice demonstrated that neutrophils not only adhere to the activated endothelium, but additionally capture circulating sickle RBCs, thereby favoring cellular interactions that culminate in the obstruction of postcapillary venules ^56,57. This may seem intuitive since neutrophils represent 50–70% of total circulating leukocytes in humans, while in mice they account for only 10–25% ^58,59. The intravital imaging data revealed, however, that over 80% of total leukocytes participating in cellular interactions during vaso-occlusion in mice are neutrophils ⁵⁷. Recent findings from in vivo lung imaging of pulmonary arterioles in SCD mice have shown the occlusion of pulmonary arterioles as a mechanism driven by neutrophil extracellular traps (NETs) and neutrophil-platelet aggregates ^60–62, supporting the involvement of neutrophils in causing vaso-occlusion of pulmonary vessels and in turn causing acute chest syndrome.

3.1. Central role of neutrophils in SCD-related vaso-occlusion

Vaso-occlusion in SCD is often initiated by inflammatory stimuli and erythrocyte-derived DAMPs released during hemolysis. Systemic vaso-occlusion preferentially occurs in small vessels, mainly postcapillary venules ⁶³, while pulmonary arterioles are the primary site of occlusion in the lungs ⁶⁰. Although RBC-endothelium interactions were initially described ⁶⁴ and for many years considered the sole or main component of vascular occlusions in SCD, leukocytes were later identified as key players in triggering these events ⁶⁵.

3.1.1. Neutrophils, RBCs and the endothelium

Neutrophil recruitment and interaction with the endothelium require adhesion molecules expressed by activated endothelial cells. The most important molecules for this step are the endothelial P- and E-selectins, which bind to glycosylated ligands on the surface of neutrophils (and other inflammatory leukocytes). P-selectin preferentially binds to P-selectin glycoprotein ligand-1 (PSGL-1), while E-selectin binds to E-selectin ligand-1 (ESL-1) and CD44 ⁶⁶. Such interactions are weak but favored by slowed blood flow during inflammation. They also activate the expression and activation of β2 integrins on neutrophils, most importantly lymphocyte function-associated antigen 1 (LFA-1, α_Lβ₂, CD11a/CD18) and macrophage antigen 1 (Mac-1, α_Mβ₂, CD11b/CD18), which bind to intercellular vascular adhesion molecules (ICAMs) expressed by endothelial cells and promote firmer interactions and crawling of rolling neutrophils with the vessel wall (Fig. 1C) ⁶⁷.

Circulating endothelial cells from individuals with SCD are abnormally activated, with exacerbated expression of ICAM-1, vascular cell adhesion molecule-1 (VCAM-1), as well as E- and P-selectin as a result of the inflammatory milieu promoted by the disease ^68,69. Consistently, in vitro culture of human umbilical vein endothelial cells (HUVECs) with TNFα, a potent inflammatory stimulus widely used to simulate SCD-related inflammation, enhances the expression of endothelial adhesion molecules and selectins ^70,71. Cytokine-induced expression of endothelial adhesion molecules and selectins in vitro has shown to be repressed by NO, possibly through inhibition of NF-kB in endothelial cells ⁷², suggesting that reduced levels of NO in SCD patients exacerbates the inflammatory response. Soluble VCAM-1, ICAM-1 and E-selectin are elevated in patients with SCD and may function as important markers of endothelial dysfunction and disease severity, since they have been associated with pulmonary hypertension, liver disease, renal dysfunction and early mortality ⁷³.

As indicated above, in vivo imaging of the microvasculature by intravital microscopy has become a fundamental tool for the understanding of the dynamics of neutrophils during inflammation. Intravital microscopy of the cremaster microcirculation in SCD mice lacking both P- and E-selectin showed a reduction in the rolling and adherence of leukocytes in postcapillary venules. These mice were also protected from vaso-occlusion induced by TNFα, compared with SCD mice with functional endothelial selectins ⁵⁶. Adherent neutrophils, which are larger and more rigid than erythrocytes, potentially delay the blood flow by increasing the RBC transit time in circulation and favoring their sickling and entrapment in small vessels. Interestingly, adherent leukocytes have been observed to capture RBCs during vaso-occlusion, especially sickle-shaped RBCs, and the number of RBC interactions per adherent leukocyte was drastically reduced in the absence of P- and E-selectins ⁵⁶. Further analysis of RBC-leukocyte adhesion in SCD mice showed that such interactions were mostly affected by E-selectin expression through upregulation of Mac-1 activity ⁷⁴. E-selectin is expressed exclusively by endothelial cells and regulates the expression of Mac-1 by neutrophils (Fig. 2) ⁷⁵. Mac-1 has been identified as the molecule responsible for the capture of adherent RBCs during vaso-occlusion in SCD mice (Fig. 1C), and adherent neutrophils from SCD mice display higher Mac-1 activity, which is notably reduced in the absence of E-selectin ⁷⁴. Mac-1 is markedly upregulated also in neutrophils from SCD patients, compared with healthy individuals ^76,77, and it has been demonstrated to mediate in vitro interactions of human neutrophils with endothelial cells ⁷¹ and matrix extracellular ligands (e.g. fibronectin) ⁷⁸.

Figure 2. Effector mechanisms of neutrophils in sickle cell disease.

Left: Neutrophil activation is characterized by increased α_Mβ₂ integrin (Mac-1) activity, which is triggered by the contact of neutrophils with the activated endothelium and the consequent interaction between E-selectin and its ligand ESL-1. Mac-1 can also be activated by heme through toll-like receptor 4 (TLR4) signaling. Activation of TLR4 by heme leads to reactive oxygen species (ROS) generation and inflammasome formation in neutrophils. Further stimulus for neutrophil activation is provided by extracellular ROS and inflammatory mediators released by activated endothelial cells, platelets and leukocytes. Right: Damage associated molecular patterns (DAMPs), e.g. heme, ROS, and inflammatory cytokines, promote neutrophil extracellular traps (NETs) formation through ROS generation and gasdermin-D (GSDMD) activation. NETs are characterized by DNA, histones, and granule contents, such as myeloperoxidase (MPO) and elastase (NE). Entrapment of platelets by NETs promote thrombi formation and aggregation between platelets and neutrophils.

3.1.2. Neutrophil-platelet aggregates

While RBCs and neutrophils are the main drivers of systemic vaso-occlusion, the involvement of platelets in adhesive interactions with the endothelium has been described with great relevance for pulmonary vaso-occlusion. Platelets from SCD patients circulate in an activate state, with elevated expression of P-selectin ⁷⁷. P-selectin expressed by platelets can bind to PSGL-1 expressed by adherent neutrophils ⁷⁹ and interactions may additionally occur between Mac-1 on neutrophils and glycoprotein Ibα (GPIbα) on platelets (Fig. 1C) ^74,80,81. Platelet activation leads to the release of granule contents, which is rich in inflammatory mediators, and microvesicles that activate endothelial cells and promote leukocyte recruitment and activation (Fig. 2). Platelets from SCD individuals are capable of promoting endothelial activation in vitro ⁸² and are enriched with IL-1β, an inflammasome-derived cytokine released through platelet extracellular vesicles and shown to further activate neutrophils, endothelial cells and platelets ⁶¹.

Neutrophil-platelet aggregates are abundant in the blood of SCD patients ^77,83 and have been identified occluding pulmonary arterioles in SCD mice challenged with intravenous bacterial lipopolysaccharide (LPS), a TLR4 ligand, or Oxy-Hb, a precursor of heme ^60,62. TLR4 activation by heme is known to prime inflammasome activation in leukocytes and to induce platelet NLRP3-inflammasome, with consequent generation of neutrophil-platelet aggregates (Fig. 2) ⁶¹. Heme-induced TLR4 signaling in endothelial cells leads to the activation of the NFkB pathway, thereby promoting endothelial activation and vaso-occlusion in SCD mice ⁸⁴. Activation of NFkB by heme has been also reported in human neutrophils in vitro and resulted in an augmented adhesiveness of neutrophils to fibronectin and endothelial ICAM-1, increased ROS generation, and enhanced Mac-1 activity ⁷⁸.

3.1.3. Neutrophil extracellular traps (NETs)

Another pathway of SCD vaso-occlusion lies on the potential of heme to trigger one of the most intriguing effector mechanisms used by neutrophils exposed to pathogens or sterile stimuli – the generation of NETs ⁸⁵. NETs are defined as extracellular structures of DNA coated with histones and granule-derived proteins, such as neutrophil elastase (NE), cathepsin G, and myeloperoxidase (MPO) ^85,86.

One of the mechanisms of NET generation is through induction of nuclear decondensation and release of cellular contents by gasdermin-D (GDSMD) (Fig. 2) ⁸⁷. Briefly, activation of neutrophils triggers ROS generation by NADPH oxidase, leading to the activation of protein-arginine deiminase 4 (PAD4), which mediates histone hypercitrullination and induces chromatin unfolding ⁸⁸. ROS-induced activity and release of MPO and NE from azurophilic granules promote degradation of cytoplasmic actin filaments and contribute to chromatin decondensation. Finally, the formation of pores in the plasma membrane and in the azurophilic granules by GSDMD leads to cell lysis and NET release ^87,89.

NETs play an important role in controlling infections; however, they can be harmful to the endothelium and induce further inflammatory responses ⁹⁰. Additionally, activated platelets or neutrophils-platelet aggregates can be entrapped into circulating NETs and promote thrombi formation ⁹¹. Growing evidence suggest that NETs favor the aggregation of neutrophils and platelets in the pulmonary arterioles, leading to pulmonary vaso-occlusion and acute lung injury ^27,62. NETs and soluble NET components have been observed in the lungs and peripheral blood, respectively, of TNFα-stimulated SCD mice. In non-SCD control mice, in which hemolysis is absent, TNFα only induced NET formation when administered with heme. Consistently, treatment of SCD mice with hemopexin, a natural scavenger of heme, abrogated NET formation ²⁷. Recent and intriguing findings showed that NETs found in the lungs of SCD mice were not locally originated, but instead originated remotely and traveled intravascularly from the liver to the lungs, where they finally promoted vaso-occlusion ⁶².

3.2. Neutrophils as a therapeutic target in SCD

The understanding of vaso-occlusion as a multicellular phenomenon, driven by different inflammatory pathways and blood cell types, expands the range of targets for the development of new therapeutic approaches to treat and prevent SCD complications. Such strategies may focus on the early instigating event by preventing HbS polymerization and subsequent events leading to vaso-occlusion, by inhibiting downstream pathways that directly promote vaso-occlusion or, indirectly, by attenuating the vicious cycle of HbS polymerization, hemolysis and inflammation. The identification of neutrophils at the initiation and propagation stages of vaso-occlusions has placed these cells in the spotlight.

Targeting pro- and anti-inflammatory pathways is an approach that has been tested in SCD mouse models and shown favorable outcomes. Neutralization of inflammasome cytokines by specific antibodies, such as anti-IL-1β and anti-IL-18, ameliorated endothelial activation, inflammation, and vaso-occlusion triggered by hypoxia or TNFα in SCD mice ^92,93. Likewise, treatment with anti-inflammatory cytokines, for example transforming growth factor beta 1 (TGF-β1), a major immune regulator, protected SCD mice from TNFα-induced vaso-occlusion and reduced the adhesion of human neutrophils in vitro ⁷⁰. Conversely, blockage of TGF-β1 signaling by an anti-TGFβ antibody aggravated vaso-occlusion in these mice ⁷⁰. Although promising, modulation of cytokines may have undesirable effects in several pathways unrelated to vaso-occlusion due to their pleiotropic action. In this scenario, more specific approaches appear to be a better and safer choice.

Crizanlizumab, a humanized P-selectin antibody, is a prominent example of therapy with direct action in the adhesive interactions between blood cells and the endothelium. A clinical trial reported a 45% reduction in the frequency of vaso-occlusive crisis experienced by SCD patients receiving anti-P-selectin therapy compared with SCD patients receiving placebo ⁴⁴. Likewise, P-selectin inhibition ameliorated lung vaso-occlusion in SCD mice by preventing the formation of large neutrophil-platelet aggregates in the pulmonary arterioles ⁶⁰. The partial benefits achieved with anti-P-selectin treatment in both human and murine studies support the concept of multiple pathways involved in SCD vaso-occlusion. In line with these findings, a recent study reported a P-selectin-independent mechanism of lung vaso-occlusion in SCD mice. They showed that GSDMD, an effector protein of inflammasome signaling and NETs release, was overexpressed in SCD human neutrophils. Combination of both strategies, i.e., P-selectin blockage and GSDMD inhibition, potentialized the protection offered by the isolated approaches and resulted in complete abolition of lung vaso-occlusion in SCD mice ⁶².

3.3. Neutrophil heterogeneity and circadian considerations

A final (and poorly explored) question when considering how neutrophils trigger vascular disease in SCD is whether different populations exist that aggravate, or conversely alleviate, inflammatory disease in these patients. Indeed, although neutrophils were classically thought to be a homogeneous population, growing evidence of phenotypic and functional heterogeneity now exists that adds another layer of complexity to the participation of these cells in the pathophysiology of SCD and other inflammatory conditions. The concept of neutrophil heterogeneity arises from observations of distinct stages of maturation, function, activation state, and expression of surface proteins found in the total neutrophil population derived from the bone marrow, blood or other tissues ⁹⁴. Whether such differences reflect a transient state, or the existence of pre-defined subpopulations is still a matter of discussion that are relevant when considering therapeutic avenues for neutrophil-driven diseases.

An interesting point of discussion is the diversity of neutrophils acquired while in the circulation (Fig. 3). These phenotypic variations of murine and human circulating neutrophils have been shown to follow circadian oscillations ^95,96. The number of circulating neutrophils is determined by the release of “younger” (also referred to as fresh) neutrophils from the bone marrow and by the clearance of senescent neutrophils from the blood by the bone marrow and other organs such as spleen and liver. Physiological clearance of neutrophils occurs at the end of the animal’s resting period (daytime for mice; night-time for humans), when they migrate back to the bone marrow or other tissues, in a chemokine-dependent manner ^8,95. The retention and migration of neutrophils in the bone marrow compartment is regulated by CXCL12, a chemokine produced by stromal cells within this organ that binds to its receptor CXCR4 on the surface of neutrophils. The expression of CXCR4 is low in fresh neutrophils but increases as they age in the circulation and instructs their homing back to the bone marrow ^95,97. In contrast, freshly released neutrophils express high levels of L-selectin (CD62L), an adhesion molecule that regulates neutrophil recruitment to peripheral tissues and loss of L-selectin by neutrophils occurs gradually as they transit in circulation ^95,98. The critical question is whether this phenomenon of circadian aging may underlie some of the pro-inflammatory activities of neutrophils in the context of SCD.

Figure 3. Sources of neutrophil heterogeneity in sickle cell disease.

The release of freshly produced neutrophils from the bone marrow occurs at the beginning of the rest period, which is marked by the onset of light in mice and darkness in humans. As these neutrophils circulate in the blood, they downregulate the expression of L-selectin and upregulate the expression of chemokine receptor 4 (CXCR4) over time. Neutrophil aging is also accompanied by increased expression and activation of α_Mβ₂ integrin (Mac-1). Clearance of aged neutrophils takes place in the bone marrow or other organs, such as liver and spleen, at the end of the resting period. Accumulation of circulating aged neutrophils in SCD mice is a result of the inflammatory milieu promoted by the disease and a response to signals generated by the gut microbiota. Microbial signals are under influence of external factors, such as antibiotics, or endogenous factors, including signals from the nervous system.

An augmented number of neutrophils expressing an aged phenotype, defined by low expression of CD62L and high expression of CXCR4, has been observed in the blood of SCD mice ⁹⁹. These neutrophils upregulate Mac-1 activity, TLRs signaling, and NETs formation, and have been associated with acute vaso-occlusion and long-term organ damage. The accumulation of aged-like neutrophils in the blood of SCD mice was prevented by chronic treatment with broad-spectrum antibiotics, suggesting that signals from the gut microbiota controls the aging process of neutrophils (Fig. 3) ⁹⁹. One possible mechanism connecting aging with the microbiome is the crosstalk between gut microbiota and the nervous system. Expansion of aged neutrophils has been observed in the blood of SCD mice subjected to a restraint protocol of psychological stress and resulted in worsening of inflammation and vaso-occlusion, which were prevented by gut microbiota depletion ¹⁰⁰. Not only the intestinal barrier is compromised in SCD, but intestinal dysbiosis has been reported in both SCD patients and mice ^101,102. Since the gut microbiota is essential for the regulation of the innate immune system, intrinsic or acquired perturbations in the composition of the gut microbiota can potentially affect circulating neutrophils in SCD.

The frequency of circulating neutrophils with high expression of CXCR4 is reportedly elevated in the blood of SCD patients ¹⁰³. These cells demonstrate higher activity of Mac-1 and propensity to form neutrophil-platelet aggregates, with a greater number of platelets bound per neutrophil, when compared to neutrophils expressing low CXCR4. Platelets have been demonstrated to induce the expression of CXCR4 on neutrophils in vitro, through the release of serotonin, which is elevated in the plasma of SCD individuals. Incubation of healthy neutrophils with purified serotonin or with SCD plasma, in the absence of platelets, also induced the expression CXCR4. Corroborating the involvement of serotonin in the phenotypic modulation of neutrophils in SCD, the use of a non-selective inhibitor of serotine receptors prevented neutrophils from expressing CXCR4 in the presence of SCD plasma ¹⁰³. Thus, an interplay between microbial or platelet derived metabolites may drive inflammatory episodes in SCD patients by modulating the activity of neutrophils.

The microenvironment in different tissues or under different stimuli can induce the acquisition of different functions by neutrophils and may underlie the acquisition of distinct phenotypes that differ in the expression of important activation markers ⁹⁴. Early intravital studies reported that less than 20% of total adherent leukocytes capture sickle RBCs during TNFα-induced vaso-occlusion in SCD mice ⁵⁶, an observation that suggests diversity in the activity of these cells during intravascular inflammation. Variations in the expression and activation of the RBC receptor Mac-1 by neutrophils in SCD, for example, has also been associated with changes in their physical density, which may in turn be related to maturation or degranulation ¹⁰⁴. The concept of different density subsets of neutrophils derives from observations that during isolation of neutrophils by a Ficoll density gradient, a small fraction of low-density-neutrophils (LDN) is found in the mononuclear cell layer. These LDN represent a heterogeneous population composed by both mature and immature cells and have been extensively studied in the context of cancer due to their pro-tumoral properties ^105,106. An elevated number of LDN is found in the blood of SCD patients and SCD mice and these cells feature higher expression and activity of Mac-1 and increased adhesion in vitro ¹⁰⁴, suggesting that LDN may be potential promoters of vaso-occlusion in SCD.

While not discussed here, other sources of neutrophil diversity emanating from the unique conditions present in these patients may dictate the pathophysiology of SCD, not only inside vessels but also in peripheral organs. For example, the observation that neutrophils control angiogenesis in lungs and intestine during early life or upon genotoxic injury ¹⁰⁷ suggests that this or other functions may also be affected in SCD patients, and this may be an attractive area of exploration in the future.

4. CONCLUDING REMARKS

The protagonist role of neutrophils in SCD pathology is now unquestionable. These cells promptly respond to inflammatory stimuli triggered by infection, hemolysis-derived DAMPs, endothelium-released mediators, and numerous factors secreted by other immune cells and injured tissues. Upregulation of adhesion molecules by activated neutrophils favors their cell-cell interactions and place these cells as central effectors and perpetuators of SCD-related vaso-occlusion. The wide range of functions displayed by neutrophils in SCD is mirrored by the identification of distinct phenotypes even in healthy individuals, which may in turn denote stages of a highly dynamic response to different stimuli and adaptation to different microenvironments ⁹⁴. Because neutrophils are cells with a very short lifetime and need to be produced constantly, the observations made in the context of SCD suggests that, in addition to the activation of neutrophils in the bloodstream discussed above, alterations in the granulopoietic process may underlie the presence of abnormally primed neutrophils, and this in turn opens a very attractive area of investigation and intervention.

SCD also provides a highly dynamic scenario where a broad spectrum of stimuli may affect neutrophils in many particular ways. The use of single-cell technologies may now offer a more unbiased approach to study the rich and complex biology of neutrophils in this context, including single cell transcriptomics and multiparametric cytometric assays that may allow us to define alterations of the SCD neutrophil baseline, or during inflammatory responses and vaso-occlusive episodes. Further, behavioral screening of single adherent neutrophils in the cremaster microvasculature of mice subjected to TNFα-induced vaso-occlusion has been recently described and emerges as a potentially powerful tool to elucidate the dynamics and functional states of neutrophils SCD. Indeed, the use of over 70 morpho-kinetic parameters in intravascular adherent neutrophils within the inflamed vasculature of otherwise healthy mice revealed at least three well-defined behaviors, which differed in shape, size, motility or proximity to the vessel wall. Most notably, certain behavioral patterns associated with pathological inflammation thereby allowing the identification of disease-driving genes ¹⁰⁸. These new technologies and other orthogonal approaches to investigating neutrophils may uncloud our understanding of how SCD modifies the inflammatory properties of neutrophils, and these in turn drive disease.

In sum, as we progressively increase our understanding of the biology of neutrophils and their alterations in SCD patients and animal models, we may soon develop strategies to effectively target the inflammatory complications seen in these patients, without compromising tissue homeostasis or anti-microbial defense. One step in this direction is the discovery that neutrophils are not phenotypically and functionally rigid but are instead highly dynamic and adaptable to different environments. We hope that harnessing such new concepts may provide real therapeutic options for millions of patients in the not-too-distant future.