Low‐Density Neutrophils in Systemic Lupus Erythematosus

Abstract

Patients with systemic lupus erythematosus (SLE) display increased numbers of immature neutrophils in the blood, but the exact role of these immature neutrophils is unclear. Neutrophils that sediment within the peripheral blood mononuclear cell fraction after density centrifugation of blood are generally defined as low‐density neutrophils (LDNs). Far beyond antimicrobial functions, LDNs are emerging as decision‐shapers during innate and adaptive immune responses. Traditionally, neutrophils have been viewed as a homogeneous population. However, the various LDN populations identified in SLE to date are heterogeneously composed of mixed populations of activated mature neutrophils and immature neutrophils at various stages of differentiation. Controversy also surrounds the role of LDNs in SLE in terms of whether they are proinflammatory or polymorphonuclear myeloid‐derived suppressor cells. It is clear that LDNs in SLE can secrete increased levels of type I interferon (IFN) and that they contribute to the cycle of inflammation and tissue damage. They readily form neutrophil extracellular traps, exposing modified autoantigens and oxidized mitochondrial DNA, which contribute to autoantibody production and type I IFN signaling, respectively. Importantly, the ability of LDNs in SLE to perform canonical neutrophil functions is polarized, based on mature CD10+ and immature CD10− neutrophils. Although this field is still relatively new, multiomic approaches have advanced our understanding of the diverse origins, phenotype, and function of LDNs in SLE. This review updates the literature on the origin and nature of LDNs, their distinctive features, and their biologic roles in the immunopathogenesis and end‐organ damage in SLE.

Traditional concepts of SLE immunopathogenesis

Systemic lupus erythematosus (SLE) is an autoimmune disease of unknown etiology. Its incidence has been constant in recent decades, with ~4–5 per 100,000 people affected each year and a female:male bias peaking at 9:1 in adulthood 1, 2. However, the severity and mortality are higher in childhood‐onset and male SLE 3. It is one of the most common systemic autoimmune diseases, with a prevalence of 40–100 per 100,000 people in the US 4, 5. Common hypotheses about SLE immunopathogenesis suggest that environmental triggers, such as infectious agents, operate in the context of genetic and epigenetic influences, resulting in aberrations in antigen presentation, lymphoid signaling, apoptosis, epitope modification, and antigen and immune complex (IC) clearance 6, 7. Ordinarily, SLE is a polygenic disease, with >50 lupus susceptibility loci now identified 8, 9. Risk and severity are also associated with an increasing number and additive effect of susceptibility alleles 10. Given that SLE is an immunologic disease, it is not surprising that the top genetic associations with disease are those associated with immunity, including the HLA genes IRF5, IRF7, IRAK1, TNFAIP3, TNIP1, IFIH1, TYK2, and C1Q 11.

Traditionally, SLE has been considered a disease of perturbed adaptive immunity, due to the critical pathogenic roles of B cells and T cells 12, 13. Characteristic antinuclear antibodies (ANAs), such as anti–double‐stranded DNA, anti‐RNA, and anti‐RNA–associated proteins, are produced by autoreactive B cells through both extrafollicular pathways and germinal center reactions 14, 15. These ANAs form ICs with nuclear material released following excessive or inappropriate cell death processes. The subsequent IC deposition in tissues results in a proinflammatory loop, which involves the recruitment of multiple immune cells with cytokine release and the destruction of multiple organs 16, 17. Type I interferons (IFNs) comprise a family of cytokines that have been associated with SLE. They are divided into 5 classes (IFNα, IFNβ, IFNε, IFNκ, and IFNω), with IFNα being further categorized into 12 subtypes 18. IFNα was first described in the peripheral blood of SLE patients by Preble et al in 1982 19. Since then, an “IFN signature” has been identified in the peripheral blood mononuclear cells (PBMCs) of SLE patients 20, 21.

This up‐regulation of IFN‐regulated genes occurs in a subset of adult SLE patients and may reflect differences in genetics, environmental triggers, or disease severity 21. It is also now clear that although IFNα subtypes are important in SLE, additional type I IFNs, such as IFNβ and IFNκ, play key roles in disease 18, 22, 23. This interest in type I IFNs initiated a plethora of clinical trials with therapeutic targets directed at IFNα and type I IFN receptor, which have more recently been redirected toward those expressing the IFN signature 24. Neutrophils have generated interest for their role in the production of type I IFNs and for their proinflammatory and antiinflammatory functions in SLE 21, 25, 26, 27. Moreover, there is a burgeoning number of reports showing that they play an integral role in disease pathogenesis.

Neutrophil fundamentals

To understand dysfunction of any cell type, it is necessary to understand the underlying biology. Neutrophils are the most abundant immune cell type in the human peripheral blood, acting as the first responders during sterile and microbial insults 28. Chemotaxis and extravasation toward the potential pathogen are mediated through the activation of a number of surface receptors, including selectins and integrins, chemokine receptors, Fcγ receptors (FcγRs), and fMLP receptors. Neutrophils target microorganisms through the generation of reactive oxygen species (ROS) via respiratory burst, the release of bactericidal enzymes through degranulation, and phagocytosis, a process that engulfs the potentially pathogenic organism 26, 29, 30. More recently, NETosis has been described as a mechanism to trap potential pathogens within a network of expelled cell contents termed neutrophil extracellular traps (NETs) 29. Neutrophils also play a key role in eliminating opsonized bacteria and in antibody‐dependent cell‐mediated cytotoxicity 31, 32. They constitutively express a low‐affinity IgG receptor, FcγRIIa (CD32A) and the glycosyl phosphatidylinositol–linked FcγRIIIb (CD16B), which serves as a coreceptor 33, 34. Upon the crosslinking of FcγR in vitro, neutrophils activate, degranulate, produce ROS, and can trigger NETosis 33. In addition, neutrophils interact with the complement pathway to ensure clearance of invading pathogens, primarily through C5a 35.

The in vivo half‐life of human circulating neutrophils is 19–90 hours. This estimate is derived from the ratio of the number of neutrophils in the blood to the number of mitotic neutrophil precursors in the bone marrow 36, 37. Consistent with this relatively short lifespan, 50% of the bone marrow is devoted to neutrophil production, releasing ~5–10 × 10¹⁰ cells per day, which increases during an infection 38, 39. This lifespan increases following exposure to cytokines or other proinflammatory agents present in infected or inflamed tissue 40, 41.

Neutrophil heterogeneity is an important feature of immune pathophysiology, and therefore strategies for assessing neutrophil subsets based on their maturation have been proposed 28, 30. Maturation from committed proliferative neutrophil precursors (pre‐neutrophils) into nonproliferative immature neutrophils and then mature neutrophils can be detected, using changes in cell surface molecules 28, 30.

Low‐density neutrophils in SLE

In 1986, Hacbarth and Kajdacsy‐Balla were the first to describe the presence of “low buoyant density neutrophils” in the peripheral blood mononuclear cell (PBMC) preparations obtained from adult SLE patients 42 (Figure 1). This preparation remains common and involves the centrifugation of whole blood, mixed with media, layered over a polysaccharide or silica gradient with a density of ~1.077 gm/ml (e.g., Ficoll‐Paque, Histopaque‐1077, Percoll) for humans 43. Due to their relative high density, neutrophils end up below the Ficoll layer, on top of the erythrocyte fraction (often termed high‐density neutrophils [HDNs]). The PBMC fraction is found in the interphase between the Ficoll layer and the plasma 43. In this original study, the authors hypothesized that humoral factors in patient plasma induced activation of the neutrophils in vitro, causing degranulation and decreasing their buoyant density, leaving them to settle with the PBMCs as low‐density neutrophils (LDNs) 42. Supporting this idea, stimulation of whole blood with fMLP was shown to lead to an increase in the percentage of LDNs at the PBMC fraction in a dose‐dependent manner, representing up to 12.5% of the population 44, 45. Furthermore, the activation of neutrophils led to increased expression of CD66b, which is mobilized from intracellular granules to the cell membrane during degranulation, possibly contributing to the LDN increase 45.

Figure 1.

Phenotypic and biologic properties of low‐density neutrophils (LDNs) in systemic lupus erythematosus (SLE). SLE LDNs are found within the peripheral blood mononuclear cell fraction. CD11b, CD16, and CD66b are common markers to identify mature high‐density neutrophils (HDNs). However, LDNs comprise a heterogeneous population of CD10− pre‐neutrophils, immature neutrophils, and CD10+ mature neutrophils. SLE LDNs secrete increased levels of proinflammatory cytokines, have impaired phagocytosis, and enhanced neutrophil extracellular trap (NET) formation via mitochondrial reactive oxygen species (mtROS) production, leading to elaboration of oxidized mitochondrial DNA (mtDNA) and interleukin‐17 (IL‐17). RBC = red blood cell; TNF = tumor necrosis factor; IFNα = interferon‐α; 8‐OHdG = 8‐hydroxydeoxyguanosine.

In 2003, Bennett et al performed microarray analysis of PBMCs from pediatric SLE patients and identified high expression of neutrophil‐specific genes 20. This “granulocyte signature” was due to an increase of LDNs in the PBMC layer 20. This signature was not caused by steroid treatment as it was demonstrated in several newly diagnosed, untreated pediatric SLE patients 20. Moreover, intravenous pulse high‐dose glucocorticoid treatment extinguished the IFN signature without affecting the granulocyte signature in pediatric SLE PBMCs 20.

Among low‐density granulocytes, which encompass neutrophils, eosinophils, and basophils 26, 43, only LDNs have been described in SLE. The frequency of LDNs found contaminating the PBMC fraction is significantly higher in SLE patients, at ~17% of total PBMCs, compared to ~5% in healthy donor PBMCs 26, 46, 47 (Table 1). Neutrophils with low‐density features have also been described in other rheumatic diseases such as antineutrophil cytoplasmic antibody–associated vasculitis, myositis, primary antiphospholipid syndrome, and psoriasis 48, 49, 50, 51, 52, 53. In summary, LDNs form a contaminating granulocytic population in the PBMC layer of fractionated blood, which is increased in SLE.

Table 1.

Biologic and clinical characteristics of LDNs in SLE^a

Prevalence, %	HD 0.81–5.00; SLE 2.37–17.00
Morphology	Promyelocytes, myelocytes, metamyelocytes, bands, and segmented neutrophils
Immunophenotype	All LDNs CD3−CD19−CD20−CD56− and CD11bhighCD14−/lowCD15+CD16+CD33+CD66bhigh; mature LDNs CD10+; immature LDNs CD10−
Functional properties	Increased TNF, IFNα, IL‐17+ NETs; decreased phagocytosis; decreased chemotactic activity; increased mtROS production; increased spontaneous NETosis with NET‐mtDNA release
Homeostatic/pathologic relevance	CD10+ LDNs associated with noncalcified plaque burden severity, vascular inflammation, and lower HDL cholesterol efflux capacity in SLE patients, and negatively correlated with renal function in white SLE patients; CD10− LDNs positively correlated with proteinuria in white SLE patients

^aLDNs = low‐density neutrophils; SLE = systemic lupus erythematosus; HD = healthy donors; TNF = tumor necrosis factor; IFNα = interferon‐α; IL‐17 = interleukin‐17; NETs = neutrophil extracellular traps; mtROS = mitochondrial reactive oxygen species; mtDNA = mitochondrial DNA; HDL = high‐density lipoprotein.

SLE LDN surface markers, nuclear morphology, and activation markers

LDNs can be distinguished from monocytes in the PBMC layer by their high granularity and expression of surface molecules using flow cytometry 20, 26, 47 (Table 1). LDNs also lack major histocompatibility complex class II and the costimulatory molecule CD86 (B7.2), in contrast to monocytes, and express CD15, CD16b (FcγIIIb), CD33, and CD11b (Table 1). Comparatively, SLE LDNs express higher levels of the activation markers CD11b, CD66b, CD63, and CD107a, relative to autologous HDNs 26, 47. CD11c, CD31, granulocyte colony‐stimulating factor (G‐CSF) receptor, and granulocyte–macrophage colony‐stimulating factor receptor have also been found on their surface, in contrast to HDNs 54, 55. However, they do not differ with regard to surface L‐selectin (CD62L), which is a sensitive measure of activation 26.

More recently, 2 subpopulations of SLE LDNs which comprise immature CD10− and mature CD10+ neutrophils have been identified 46, 56 (Table 1). Similarly, morphologic analysis of SLE LDNs shows all stages of neutrophil development, including promyelocytes, myelocytes, metamyelocytes, bands, and segmented neutrophils 20, 46, 54, 56 (Table 1). Moreover, up to 60% of human SLE LDNs are mature, in contrast to HDNs, which consist of >90% mature cells 26, 54.

Genetics, epigenetics, and gene expression of SLE LDNs

Analyses of genomic instability have shown increased copy number alterations and microsatellite instability in SLE LDNs compared to matched HDNs from the same individuals 8. These somatic alterations are consistent with the notion of DNA strand break repair and a replication error–prone status, supporting a model whereby genomic damage contributes to the development of an abnormal population of neutrophils 8. Consistent with these findings, chronically elevated ROS, which often occurs in SLE, affects several aspects of DNA damage response and may lead to accumulation of such genomic changes 57, 58. An assessment of epigenetic accessibility by Coit et al also revealed a robust and consistent demethylation of IFN signature genes in SLE LDNs and HDNs compared to healthy donor HDNs 59.

A more stratified transposase‐accessible chromatin sequencing study by Mistry et al has shown that immature SLE LDNs have more open peaks compared to mature SLE LDNs, reflecting enhanced chromatin accessibility and suggesting increased gene activity 56. Similarly, an analysis of the transcriptome of SLE LDNs using bulk RNA sequencing showed increases in cell cycle progression genes in the CD10− population, confirming their immaturity 56. SLE LDNs also express higher messenger RNA (mRNA) levels for >200 genes relative to autologous or healthy donor HDNs, including serine proteases, bactericidal proteins, and other molecules involved in neutrophil regulation of inflammatory responses 60. Interestingly, type I IFN signaling transcripts, which include MX1, IFIT3, IFI44, and RSAD2, are increased only in mature CD10+ SLE LDNs and not in the immature CD10− population 54, 56. However, this may simply reflect the exposition time to the inflammatory milieu. Taken together, these data suggest that LDNs may be a mixed cell population consisting of mature cells, together with immature and activated/regulatory populations.

Cytokine profiles of SLE LDNs

Resting healthy donor HDNs, SLE HDNs, and LDNs show similar levels of cytokine release 26. However, phorbol myristate acetate–stimulated SLE LDNs secrete more tumor necrosis factor (TNF) and have higher levels of IFNα mRNA compared to autologous SLE and healthy donor HDNs 26 (Table 1). Furthermore, cultured supernatants from resting or stimulated SLE LDNs induce type I IFN–inducible genes, in contrast to HDNs from SLE patients or healthy donors, suggesting that they release Toll‐like receptor (TLR) ligands and/or IFNα itself 26. Although plasmacytoid dendritic cells provide the highest levels of IFNα upon stimulation, all other nucleated cells have the capacity to produce IFNα 25, 61. SLE LDN supernatants also induce IFNγ, TNF, and lymphotoxin α (LTα) from T cells, suggesting a proinflammatory phenotype of LDNs 47. However, more research is required to fully comprehend the capacity of LDNs in cytokine production.

SLE LDNs show a propensity for NETosis rather than phagocytosis

LDNs from SLE patients have a lower capacity for phagocytosis compared to autologous HDNs from SLE patients or healthy donors 26 (Table 1). This may be due to the relatively high proportion of immature cells, since CD10− SLE LDNs are less effective at phagocytosis compared to mature CD10+ SLE LDNs 56. Findings from a study by Lood et al suggest that this reduction may be due to an increase in the cleavage of FcγRIIa following activation by TLR‐7/8 ligands 62. In contrast, SLE LDNs show an increase in spontaneous NETosis and the release of self‐reactive material compared to autologous and healthy donor HDNs 27, 60, 63. Enhanced NETosis also results in comparatively higher secretions of interleukin‐17 (IL‐17), an important cytokine associated with T cell activation, particularly in autoimmunity 60, 64 (Table 1). The NETosis in SLE LDNs is, in part, due to an increase in mitochondrial ROS (mtROS) release 27, 65 (Figure 1 and Table 1). Importantly, mtROS production is sufficient for the generation of NETs even in the absence of functional NADPH oxidase 27. The oxidized mitochondrial DNA (mtDNA) derived from the LDN NETs is proinflammatory, activating stimulator of IFN genes to induce type I IFN signaling in target cells 27. Interestingly, mitochondrial stress alone can induce the release of mtDNA and induce IFN signaling. Inhibition of this process decreases NET formation and mtDNA release, reducing lupus‐like disease in a model system 66.

Additional cytokines, ligands, and pathways associated with NETosis have also been described in the pathogenesis of SLE. Lipopolysaccharide was described early on in the literature as a major stimulant of NETosis. However, in the absence of a bacterial infection, the specific stimulus in SLE remains unclear. IL‐18 has been shown as an effective stimulus of NET release by SLE LDNs in active lupus nephritis 67, 68. Lood et al have also demonstrated that RNP‐containing ICs can induce mitochondria mobilization, mtROS production, and the release of oxidized mtDNA 27. Similarly, van Dam et al demonstrated that SLE ICs are able to induce nonlytic NETosis in healthy donor HDNs via FcγR signaling, with rapid extrusion of NETs enriched with oxidized mtDNA 69. Therefore, ICs, mitochondrial dysfunction, and cytokines may all contribute to increased NETosis in SLE LDNs.

SLE LDNs and myeloid‐derived suppressor cells are distinct cell types

In 2007, Gabrilovic et al suggested the term “myeloid‐derived suppressor cells” (MDSCs), informed by the myeloid origin, immunosuppressive function, and systemic expansion of these cells in a cancer‐related context 70. MDSCs consist of 2 large groups of cells: polymorphonuclear MDSCs (PMN‐MDSCs) and monocytic MDSCs (M‐MDSCs) 71. PMN‐MDSCs are phenotypically and morphologically similar to neutrophils, while M‐MDSCs are similar to monocytes 71. Traditionally, PMN‐MDSCs are enriched in the low‐density PBMC fraction using density centrifugation methods, in a similar manner to LDNs 71. PMN‐MDSCs and SLE LDNs also express similar surface molecules, including CD11b+CD14−CD15+CD66b+, leading to a hypothesis that they are the same type of cells 26, 46, 70 (Table 2). However, an evaluation of the literature on PMN‐MDSCs, primarily from cancer research, and SLE LDNs also reveals differences between the neutrophil groups (Table 2). The lectin‐like oxidized low‐density lipoprotein receptor 1 was identified to distinguish PMN‐MDSCs from other neutrophils, although it is expressed on only one‐third of PMN‐MDSCs in cancer patients 72. Furthermore, it is expressed on LDNs from healthy donors and SLE patients at similar levels 47, eliminating its role in categorization.

Table 2.

Phenotypic, biochemical, molecular, and functional properties of SLE LDNs and PMN‐MDSCs^a

	SLE LDNs	PMN‐MDSCs
Physical characteristic
Density	Low	Low
Morphology	Promyelocytes, myelocytes, metamyelocytes, bands and segmented neutrophils	Metamyelocytes, bands, and segmented neutrophils
Surface marker
FSC	NA	High
CD11b	+++	++
CD14	−/+	−
CD15	+++	+++
CD66b	+++	+++
LOX‐1	+/++/+++	+
Biomarker
ROS	+	+++
ARG1	++	++
PD‐L1	+	+
Immunometabolic status
ER stress	++	++
Functional test
Inhibition of T cell proliferation	No	Yes
Inhibition of IFNγ production	No	Yes

^aPositive and negative signs indicate the level of expression of relevant markers in SLE LDNs or polymorphonuclear myeloid‐derived suppressor cells (PMN‐MDSCs). NA = not applicable; LOX‐1 = lectin‐like oxidized low‐density lipoprotein receptor 1; ROS = reactive oxygen species; ARG1 = arginase 1; PD‐L1 = programmed death ligand 1; ER = endoplasmic reticulum (see Table 1 for other definitions).

The hallmark of PMN‐MDSCs is their ability to suppress T cell function 72, which is contrary to findings from LDN studies in SLE (Table 2). Cultured SLE LDNs, or their supernatants, fail to suppress proliferation of activated healthy donor naive CD4+ T cells 47. Moreover, the addition of SLE LDNs enhances the production of IFNγ, TNF, and LTα by stimulated CD4+ T cells, in contrast to HDNs 47. This may be due to the comparative heterogeneous nature of SLE LDNs discussed above. The addition of predominantly immature CD10− SLE LDNs enhances T cell proliferation in vitro; however, isolated CD10+ LDNs can promote immunosuppression via a CD18‐mediated contact‐dependent arginase 1 release 46. Similarly, studies on immature CD10− LDNs from donors receiving recombinant human G‐CSF show increased T cell proliferation and IFNγ production 46, suggesting that maturity is a key factor in suppressive function. In contrast to SLE LDNs, morphologic examinations of PMN‐MDSCs in cancer patients showed segmented cells, and therefore maturity, in 8 of 9 studies 43. It is currently unknown whether LDNs from pediatric SLE patients are suppressive or proinflammatory.

Programmed death ligand 1 (PD‐L1) is expressed in MDSCs, and this may contribute to the success of recent anti–PD‐L1 immunotherapies used in cancer 70 (Table 2). Inhibition of this suppressive receptor, particularly on T cells, results in increased activation of the immune system and elimination of tumor cells. On MDSCs, inhibition of PD‐L1 may reduce their suppressive action. Interestingly, PD‐L1 has also been detected in LDNs from SLE patients 47. Further research is necessary to determine whether activating PD‐L1 will be an effective novel therapy in SLE through its suppression of the immune system.

LDNs are associated with organ damage in SLE

LDNs and premature cardiovascular disease

Cardiovascular disease (CVD) accounts for more than one‐third of all deaths in SLE patients 73. This is due to a plethora of factors, including autoantibodies, systemic inflammation, and endothelial injury, in addition to traditional CVD risk factors 74. A longer disease duration, higher damage index, and less aggressive immunosuppression are associated with increased heart disease, suggesting that immune dysregulation contributes to plaque progression and vascular complications, which are essential factors associated with CVD 73.

Endothelial dysfunction is an early central phase in the evolution of atherosclerosis and has been described as the intermediate link between CVD risk factors and the development of atherosclerosis 75. SLE LDNs have been associated with endothelial damage and abnormal endothelial proliferation in vitro 54. Furthermore, the reduced ability of SLE endothelial progenitor cells to differentiate into mature endothelial cells is mediated in part by IFNα produced by CD10+ SLE LDNs 26. Mature CD10+ LDNs have also been associated with noncalcified plaque burden severity and lower high‐density lipoprotein (HDL) cholesterol efflux capacity in SLE patients 56, 76 (Table 1). HDL cholesterol efflux capacity is the ability of HDL to promote efflux of cholesterol from macrophages and has a strong inverse correlation with CVD 77. In support of these findings, SLE LDNs have higher levels of genes associated with the regulation of vascular inflammation and noncalcified plaque burden, including AZU1, MPO, CTSG, PRTN3, ELANE, and DEFA3, compared to healthy donors 76 (Table 1).

In vitro culture of SLE LDNs with human umbilical vein endothelial cells results in increased endothelial cytotoxicity compared to autologous or healthy donor HDNs 26, 60. This cytotoxicity is dependent on NET formation and the externalization and activation of matrix metalloproteinases 26, 60. In support of these findings, administration of a potent antioxidant and ROS inhibitor, idebenone, inhibits NET formation and improves endothelial function in the MRL/lpr murine model of SLE 65. Collectively, these data suggest that LDNs contribute to the excess cardiovascular risk in SLE and that targeting mitochondrial dysfunction or LDNs directly may be a potential treatment strategy.

LDNs, cutaneous lupus, and lupus nephritis

Netting neutrophils have been observed in biopsy samples from patients with cutaneous LE and lupus nephritis 60, 78, 79. The majority of SLE patients with elevated levels of LDNs display clinical skin manifestations, including vasculitis 26. Furthermore, skin biopsies from SLE patients with several forms of cutaneous LE revealed neutrophil infiltration and the presence of NET‐DNA 60. Similarly, lupus nephritis has also been associated with impaired NET degradation, and NET‐DNA has been observed in kidney biopsy samples from lupus nephritis patients 80. The presence of NET‐mtDNA may also suggest increased LDN activity specifically 60, 79. Given the importance of the findings of these few studies, further investigations into the functional role of neutrophils in SLE are required.

Therapeutic targeting of LDNs in SLE

LDNs represent a novel therapeutic target in SLE. Identification of the pathogenic LDN subset may facilitate development of drugs that target selective neutrophil populations while preserving critical aspects of neutrophil‐mediated host defense 81. Medications that target key events in NET formation in LDNs or promote NET clearance may provide novel therapeutic strategies in SLE. These include medications already in use for treatment of SLE, as well as novel agents under investigation (Table 3).

Table 3.

Potential therapies targeting LDNs or NET formation in SLE^a

Drug	Mechanism of action	Effect on neutrophils/NETs	Effect on disease/clinical use
Chloroquine/hydroxychloroquine	Unknown	Chloroquine inhibits NETosis in SLE LDNs in vitro	Antimalarials used as first‐line treatment in SLE
Colchicine	Possibly via inhibition of tubulin polymerization	Inhibits spontaneous NETosis in Behçet's syndrome HDNs in vitro	Used to treat SLE pericarditis
Cyclosporine	Modulates calcium‐dependent signal transduction by calcineurin inhibition	Inhibits ionomycin‐ and IL‐8–induced NETosis in healthy donor HDNs in vitro	Used to treat membranous lupus nephritis
DNase I	Enzymatic degradation of DNA	Enzymatic degradation of NET‐DNA	Well tolerated in phase I study in 17 patients with lupus nephritis
Eculizumab	Monoclonal antibody against complement C5 to inhibit the cleavage of C5 to C5a and C5b, possibly via reduction of C5a‐primed neutrophils for NETosis	ANCA‐induced NETosis in C5a‐primed healthy donor HDNs in vitro	Used to treat SLE patients with thrombotic microangiopathy
Idebenone	Antioxidant that protects cells against ROS toxicity, improves mitochondrial physiology	Inhibits spontaneous NETosis in SLE LDNs but not SLE HDNs in vitro	Reduced disease activity and organ damage in lupus mouse models
Metformin	Unknown	Decreases NET‐DNA and NET‐mtDNA from healthy donor HDNs in vitro	Open‐label study showed reduced SLE flares and steroid‐sparing effect
N‐acetylcysteine	Free radical scavenger	Inhibits NETosis and free radical formation from healthy donor HDNs in vitro	Well tolerated in phase I study in 36 SLE patients, with reduction in ADHD Self‐Report Scale scores
Tofacitinib	JAK1 and JAK3 inhibitor	Decreases NETosis in bone marrow HDNs obtained from MLR/lpr mice treated with tofacitinib	Well tolerated in a phase Ib/IIa study in 30 SLE patients, with reduction in circulating LDNs in tofacitinib‐treated group
Rituximab (followed by belimumab)	Rituximab: monoclonal antibody against CD20; belimumab: monoclonal antibody against BAFF	Decreases spontaneous NETosis in SLE HDNs ex vivo by reducing autoantibodies	Rituximab followed by belimumab was safe in a phase II study, with clinical responses in patients with severe refractory SLE
Vitamin D	Unknown	1,25(OH)2D3 decreases NETosis in SLE HDNs in vitro	Meta‐analysis of RCTs showed that vitamin D supplementation in SLE is safe and may improve fatigue

^aHDNs = high‐density neutrophils; ANCA = antineutrophil cytoplasmic antibody; ADHD = attention deficit hyperactivity disorder; RCTs = randomized controlled trials (see Table 1 for other definitions).

Conclusions

Neutrophils play a fundamental role in protection against invading pathogens, but overactivity is associated with prolonged inflammation and tissue destruction. The discovery that they fractionate with the PBMC layer in SLE revolutionized concepts pertaining to their role in disease. The neutrophil‐related gene signatures in PBMCs have been associated with common comorbidities of SLE, including cardiovascular disease and lupus nephritis 76, 82, 83, 84, suggesting their critical role in pathogenesis. The physical reasons as to why the LDNs fractionate in this layer remains a target of investigation. The parallel emergence of LDNs and PMN‐MDSCs has complicated the field, given their similar phenotypes and densities 85. However, their different functional roles in immune suppression and activation demonstrate that they are not the same.

So far, findings indicate that LDNs are a heterogeneous population, often with a large proportion of immature cells. There are no definitive markers to identify this population for in‐depth characterization. Therefore, isolating LDNs remains a multistep and lengthy process, which limits what can be done because of the rapid apoptosis occurring ex vivo. An emerging area of research will seek to understand the interplay of LDNs, NETosis, and the adaptive immune system, based on existing data showing that LDNs can regulate T cell responses 16. While much has yet to be characterized, targeting LDNs in SLE remains a promising area for novel therapeutic intervention.

AUTHOR CONTRIBUTIONS

All authors drafted the article, revised it critically for important intellectual content, and approved the final version to be published.