Neutrophil sub-types in maintaining immune homeostasis during steady state, infections and sterile inflammation

Introduction Neutrophils are component of innate immune system and a) eliminate pathogens b) maintain immune homeostasis by regulating other immune cells and c) contribute to the resolution of inflammation. Neutrophil mediated inflammation has been described in pathogenesis of various diseases. This indicates neutrophils do not represent homogeneous population but perform multiple functions through confined subsets. Hence, in the present review we summarize various studies describing the heterogeneous nature of neutrophils and associated functions during steady state and pathological conditions. Methodology We performed extensive literature review with key words ‘Neutrophil subpopulations’ ‘Neutrophil subsets’, Neutrophil and infections’, ‘Neutrophil and metabolic disorders’, ‘Neutrophil heterogeneity’ in PUBMED. Results Neutrophil subtypes are characterized based on buoyancy, cell surface markers, localization and maturity. Recent advances in high throughput technologies indicate the existence of functionally diverse subsets of neutrophils in bone marrow, blood and tissues in both steady state and pathological conditions. Further, we found proportions of these subsets significantly vary in pathological conditions. Interestingly, stimulus specific activation of signalling pathways in neutrophils have been demonstrated. Conclusion Neutrophil sub-populations differ among diseases and hence, mechanisms regulating formation, sustenance, proportions and functions of these sub-types vary between physiological and pathological conditions. Hence, mechanistic insights of neutrophil subsets in disease specific manner may facilitate development of neutrophil-targeted therapies.


Introduction
Neutrophils are one of the important components of the innate immune system and play a crucial role in protecting the host against pathogens. Neutrophils express germline encoded pattern recognition receptors (PRRs) to recognize pathogens and lead to the activation of effector functions such as production of reactive oxygen species, degranulation, phagocytosis and formation of extracellular traps [1]. In response to circadian rhythms, anatomical location, infections, sterile inflammation and ageing, neutrophils adapt themselves to physiological or pathological roles and drive various biological functions [2][3][4][5]. Protozoans, fungi, bacteria and viruses stimulate neutrophils to expel their DNA along with granular proteins to form web like structures referred as extracellular traps (NETs) [6,7]. High concentrations of antimicrobial effectors within these DNA lattices serve as a platform to activate pro-inflammatory mediators, immobilize and kill the pathogens, and simultaneously clear the infections. Aged and dead neutrophils along with accumulated NETs in inflamed tissues are cleared by macrophages and facilitate the resolution of inflammation [8]. Failure in neutrophil apoptosis and impaired clearance of dead neutrophils along with NETs components results in host tissue damage and release of pro-inflammatory cytokines [9][10][11][12]. Neutrophils also modulate functions of other cell types including vascular endothelial cells to induce pathological angiogenesis [13] and adaptive immune responses [14][15][16][17]. Accordingly, neutrophil dysfunction affects host pathophysiology and subsequently plays a role in 1 3 the pathogenesis of diseases associated with sterile inflammation such as Type 2 diabetes (T2D), autoimmune diseases, vascular disorders, digestive disorders and a variety of malignancies [4,10,[18][19][20][21][22][23].
Neutrophils were traditionally assumed to represent the homogenous population of terminally differentiated cells with definite functions. However, mounting evidence indicate the existence of neutrophil subtypes based on buoyancy, cell surface markers, localization and maturity [1]. Using pre-clinical and clinical models, studies have demonstrated functional relevance of neutrophil sub-types in both physiological conditions and pathologies associated sterile inflammation and infections. Khoyratty et al. employing RNAseq and chromatin profiling in neutrophils during their development, activation, tissue distribution and acute infection identified distinctly activated transcription factor networks indicating existence of neutrophil heterogeneity. Authors showed that KLF6 and RUNX1 were necessary for neutrophil maturation and survival; RELB, IRF5, and JUNB transcription factors were responsible for cytokine generation; ROS production and NETosis required RELB and JUNB, and phagocytosis was dependent on IRF5 and JUNB. [24]. Ballesteros et al. showed existence of tissue-specific neutrophil phenotypes based on single cell RNA and ATACseq analysis [25]. Neutrophil subtypes characterized based on buoyancy referred as low-density neutrophils (LDNs) have been demonstrated for their role in pathogenesis of various diseases including infections caused by HIV-1 and SARS-CoV-2, serum erythematous lupus and breast cancer [17,[26][27][28]. Accordingly, cell surface markers such as olfactomedin, CD177, CXCR4 + CD62L − and many others have also been basis of defining neutrophil subtypes which showed functional alterations in diseases [29][30][31]. Moreover, studies from many laboratories, including ours, have shown activation of stimulus-specific signalling pathways in neutrophils [32,33]. For example, high glucose, LPS and homocysteine representing stimuli for diabetes, infections and thrombosis, respectively, induced distinct set of kinases which were associated with specific functions of neutrophils [33]. This indicates the existence of neutrophil subtypes with confined functions in patho(physiological) conditions and further, molecular mechanisms regulating the formation, sustenance and functions of these subsets also may vary among the diseases. Moreover, these tissue-to-tissue differences and their proportions may alter in pathological conditions. Hence, mechanistic insights of neutrophil subsets in disease-specific manner may facilitate development of neutrophil targeted therapies.
In the present review, we summarize various seminal studies describing the heterogeneous nature of neutrophils during steady-state and pathological conditions. Further, we discuss challenges and therapeutic opportunities for the management of neutrophil centered diseases.

Neutrophil homeostasis during normal physiological conditions and infections
Neutrophils develop from CD34 + hematopoietic stem cells and are terminally differentiated, small granulocytes with a remarkable short lifespan [34]. Neutrophil homeostasis is maintained at different phases of life cycle including production, trafficking and clearance (Fig. 1). Exogenous and endogenous factors that determine whether a hematopoietic stem cell (HSC) in bone marrow develops into a common lymphoid progenitor (CLP) or common myeloid progenitor (CMP) cell is not clearly understood [35]. While CLP precursors develop into either T cells, B cells or NK cells, CMP develops into a granulocyte monocyte progenitor (GMP) or a megakaryocyte erythroid progenitor (MEP) [36]. These GMPs commit to neutrophil formation by transitioning into myeloblasts under the control of granulocyte colony-stimulating factor (G-CSF). Subsequently, myeloblasts mature and pass through the phases of promyelocyte, myelocyte, metamyelocyte, band cell and finally to mature neutrophils (reviewed by Vietinghoff and Ley) [37]. The immature pools are generated by the differentiation of neutrophil precursor cells that differ from mature neutrophils in terms of morphology of nucleus, expression of granular proteins, proliferative potential and enhanced transcriptional activity [38]. Immature neutrophils in humans express CD15 and CD11b, followed by an increase in the expression of CD16 and CD10 as they mature [39]. In mouse, immature neutrophils are defined by the expression of CD11b + Ly6G low Ly6B int CD117 + CD115 − and maturation begins with the elevated expression of Ly6G and decrease in the expression of progenitor cell marker c-Kit (CD117) [38]. Differential gene expression analysis in mature and immature neutrophils isolated from mouse bone marrow, blood and orthotopic pancreatic tumor showed elevated expression of cd101, which codes for a surface protein [40]. The increased expression of CD101 was observed in mature neutrophils in both bone marrow and peripheral blood. Based on the expression of CD101, authors characterized mature neutrophils with the expression of Ly6G + CXCR2 + CD101 + and immature neutrophils as Ly6G low CXCR2 − CD101 − . Further, mouse with pancreatic tumors exhibited surge in immature neutrophil pool which might serve as a biomarker for disease progression [40]. Transcription factors such as CCAAT/enhancer binding proteins (C/EBPs), Runx1 and PU.1 regulate the path of neutrophil lineage [41][42][43]. The retention of neutrophils in the bone marrow depends on the expression of the chemokine CXCL12 by perivascular cells and osteoblasts, which is a ligand for the neutrophil cell membrane chemokine receptor CXCR4 [44]. As the neutrophils mature in bone marrow, the proportion of the chemokine CXCL2 and its cell membrane receptor CXCR2 expression increases while the levels of CXCR4 decrease. This leads to the release of neutrophils from the bone marrow to circulation [44].
Neutrophils are released from the bone marrow in a regulated fashion to maintain homeostatic levels and further, increase their number in response to stress, including infections. Circulating levels of neutrophils in physiological conditions are maintained by G-CSF [45]. Semerad et al. showed that nonredundant G-CSFR signals play an important role in regulating neutrophil release from the bone marrow and maintaining physiological levels of neutrophils in the blood [46]. However, neutrophil release during infections was independent of G-CSFR signalling [46]. G-CSF potently inhibits osteoblast activity resulting in decreased CXCL12 expression in the bone marrow and involved in regulation of the release of neutrophils into the blood stream [47]. ELR + chemokines such as CXCL1-CXCL3 and CXCL5-CXCL8 produced by bone marrow-endothelial cells and osteoblasts signals through CXCR1 and CXCR2 and opposes CXCR4/CXCL12 axis [44]. Eash et al. showed balance between these pathways directs neutrophils towards bone marrow vascular space for trafficking [44]. Sympathetic nervous system provides adrenaline signals to stromal cells through β3-adrenergic receptors in bone marrow to generate circadian rhythm and regulate the expression of cxcl12 [48]. Circulating HSCs were elevated at 5 h after the initiation of day light and inversely correlated with the CXCL12 levels in BM and the release of neutrophils during the same period coincided with the levels of CXCL12 [48,49]. Upon release into circulation, neutrophils also infiltrate abundantly into the lungs, spleen and bone marrow and however studies have demonstrated presence of neutrophils in the liver, intestine, white adipose tissue and skeletal muscles [45,50].
Further under homeostatic conditions, aged neutrophils undergo clearance in different tissues. Aged neutrophils expressing decreased levels of CXCR2 and elevated CXCR4 expression leads to their migration towards bone marrow in response to CXCL12 (Fig. 1) [51]. Furze and Rankin using 111 In-labelled aged neutrophils, demonstrated that ∼32% of neutrophils from the circulation were cleared in bone Fig. 1 Life cycle of neutrophils in steady state. HSCs through series of differentiation forms promyelocyte, myelocyte, metamyelocyte, band cell and finally mature neutrophil. Steady increase in CXCR2 level and decrease in CXCR4 expression leads to release of neutrophils from the bone marrow to circulation. From circulation, neutro-phils enter into various tissues majorly into lungs, liver and spleen. As neutrophils age, increase in the expression of CXCR4, migrates back into bone marrow for clearance by resident stromal macrophages. Neutrophils are also cleared in liver and spleen 1 3 marrow in CXCR4 dependent manner and phagocytosed by resident stromal macrophages. Further, authors demonstrated approximately 29% and 26% of neutrophils were cleared in the spleen and liver, respectively [45]. Clearance of neutrophils in the liver was also observed in rat models of endotoxemia which was accompanied by an increase in P-selectin in hepatic sinusoids which led to phagocytosis of neutrophils by Kupffer cells [52].
Regulation of neutrophil production is classified into two types: steady state and emergency granulopoiesis. Changes in the microenvironment by external stimuli such as infections shift between these two stages. In steady-state granulopoiesis, ingestion of neutrophils by macrophages stimulates the activation of C/EBP-α in turn decreasing the production of cytokines and lowering the G-CSF level [53]. Whereas emergency granulopoiesis is associated with the excessive release of mature and immature neutrophils and concomitantly increase the proportion of immature neutrophils in circulation. Infections induce C/EBP-β expression and elevate cytokine levels of granulocyte-macrophage colony-stimulating factor (GM-CSF), G-CSF, Interleukin (IL) -1β and tumor necrosis factor (TNF) -α [54]. Interestingly, C/EBP-α and C/EBP-β both share common molecular interaction which are associated with granulopoiesis but differs only in the regulation of the cell cycle [55]. Emergency granulopoiesis is associated with elevated expression of chemokines such as macrophage inflammatory proteins (MIP), keratinocyte chemoattractant (KC), TNF-α and G-CSF that escalate the production of ROS in NADPH oxidase-dependent manner [54,56]. Oxidation and deactivation of phosphate and tensin homologue (PTEN) in the myeloid cell are increased by ROS activity [57]. This leads to elevated production of G-CSF and activates emergency granulopoiesis [57].

Heterogeneous nature of neutrophils during normal physiological state
Neutrophil subsets are observed in both physiological and pathological conditions [58]. Neutrophils exists in three different pools (a) proliferative, (b) circulating and (c) marginated ( Fig. 1) and the percentage of each pool of cells is influenced by maturational development and an individual's health [45]. RNA, protein and chromatin studies of mouse neutrophils from various anatomical regions including the bone marrow, gut, lung, spleen, skin and blood revealed heterogeneous properties of neutrophils [25,59]. Balleteros et al., using single cell RNA sequence analysis demonstrated neutrophils acquired distinct transcriptional profiles and phenotype in tissue specific manner. Authors showed distinct transcriptional clusters of blood, lung, spleen and bone marrow infiltrated neutrophils whereas intestine and skin derived neutrophils clustered together. This indicated phenotypic heterogeneity of neutrophils across tissues in normal steady state [25]. On similar lines, Xie et al. identified five distinct clusters of neutrophils during developmental stage represented as G0, G1, G2, G3, G4 corresponding to BM-GMP, proNeu, preNeu, immature Neu and mature Neu, respectively. Authors also identified three transcriptionally distinct subpopulation of mature neutrophils in peripheral blood represented as G5a, G5b, and G5c where G5a expressed higher levels of Mmp8 and S100a8 responsible for neutrophil migration and inflammatory response [59]. G5b neutrophils expressed a set of interferon-stimulated genes such as Ifit3 and Isg15 may be primed to combat infections. Upon experimental bacterial infection, G0 and G1 showed elevated ROS levels indicating that these early progenitor cells were primed for immune adaptation. Also, G4 and G5 neutrophils displayed upregulation of genes responsible for secretion and cytokine production [59]. Dinh et al. using mass cytometry identified earliest neutrophil progenitor cells characterized by the expression of Lin − CD117 + CD71 + CD66b + in human bone marrow neutrophil lineage [60]. Using single cell RNA sequencing methods, Wigerblad et al. identified four transcriptional clusters of neutrophils in human peripheral blood. Authors referred these clusters as Nh0 (transcriptome closely related to immature neutrophils); Nh1 (transitional phenotype); Nh2 (transcriptionally inactive) and Nh3 (enriched with transcripts related to type I IFN -inducible genes). These subsets were characterized by the expression of distinct set of transcription factors [61]. Evrard et al. identified committed proliferative neutrophil precursor (preNeu) in mice which differentiated into non-proliferating immature neutrophils and mature neutrophils. PreNeu cells ensured the production of neutrophils in homeostasis and stress [40]. Olsson et al. identified two intermediates of Gfi1-GMP positive populations (GG2, GG3) in Gfi1-GFP reporter mice. Among these two GG2 with high expression of Gfi1 represented granulocytic progenitors while GG3 expressed the highest level of Irf8 represented monocytic progenitor cells [62]. Using single cell RNA sequencing Huang et al. identified five distinct subgroups (G3, G4, G5a, G5b, G5c) of neutrophils in healthy and subjects with burnt wounds. These subsets were confined to distinct functions and showed significant alterations in transcriptome between healthy and burn conditions [63].
Several studies have demonstrated neutrophil subtypes based on expression of cell surface proteins in both physiological and pathological conditions (Table 1). Under normal physiological state, proportion of neutrophils expressed NB-1 antigen (CD177) and upon stimulation of neutrophils with fMLP although NB-1 expression was increased and the quantity of NB-1 expressing cells did not alter [29]. Olfactomedin 4 (OLFM4), a glycoprotein was expressed only in 50% of human peripheral neutrophils Shaul et al. [96] however, both subsets of neutrophils with or without expression of OLFM4 participated in all neutrophil functions such as phagocytosis, degranulation, chemotaxis and NETs [30] (Fig. 4) (Fig. 2). These LDNs isolated from healthy volunteers produced increased ROS in response to Phorbol 12-myristate 13-acetate (PMA) and showed higher phagocytic capacity compared to normal neutrophils, however, ability of NETs formation remained unaltered [65]. Aged neutrophils in circulation also adds up another layer for heterogeneity and interestingly, aged neutrophils vary in their ability to form NETs. In murine models, aged neutrophils expressing high levels of CXCR4 and decreased levels of CD62L represents an active subset exhibiting increased expression of α M β 2 integrin and NETs formation during inflammation [31]. Microbiota drives neutrophil ageing via Toll-like receptor and myeloid differentiation factor 88-mediated signalling pathways. Depletion of microbiota led to decrease in circulating aged neutrophils and improved the organ damage in endotoxin-induced septic shock mice model [31]. Peng et al. showed that in experimental metastatic cancer models, CXCR4 hi CD62L lo aged neutrophils displayed increased NETs formation robustly and promoted metastasis [66]. However, proteomics analysis in normal physiological conditions, aged neutrophils showed progressive loss of granular proteins which was associated with decreased ability to form the NETs [67]. Authors exploring endotoxin induced sepsis and acute lung infection (ALI) mouse models demonstrated diurnal differences in NETs forming ability. In ALI models, during night time, neutrophils with abundant granules formed NETs compared to day time which showed progressive loss of granules representing aged counterparts [67]. These studies indicated, ability of aged neutrophils to form NETs may depend on patho(physiological) states. Taken together, above-described studies indicate existence of neutrophil heterogeneity based on differential RNA expression, cell surface markers, density and ageing in normal physiological states and displayed an ability to perform distinct biological functions.

Neutrophil subtypes and functional relevance during host-pathogen interactions
Immediate recruitment of neutrophils to the infection/ inflammatory site is orchestrated by various chemokines such as interferon-gamma (IFN-γ), IL-8, leukotriene and C5a complement [68]. Neutrophils from healthy human individuals upon co-culture with methicillin-resistant Staphylococcus aureus (MRSA) on tissue-like scaffolds showed heterogeneous populations with differential responses in their antimicrobial activity [68]. Neutrophils displaying CD16 bright /CD62L dim hypersegmented phenotype performed normal phagocytosis but the ability to contain bacteria intracellularly was decreased, on the other hand, CD16 dim -banded neutrophil subsets effectively restricted MRSA [68]. Tsuda et al. identified two distinct subpopulations of neutrophils in mouse which were referred as PMN-1 and PMN-2 in the context of MRSA infection. PMN-1 exhibited CD49d high CD11b low (Fig. 4) phenotype with increased expression of Toll like receptor (TLR) 2, TLR5, TLR4, TLR8, and IL-2 whereas PMN-2 cells displayed CD11b high CD49d low with enhanced expression of TLR9, TLR2, TLR7, TLR4 and IL-10. PMN-1 was derived from MRSA resistant hosts, while PMN 2 was obtained from MRSA sensitive hosts [69] Authors suggested that immunocompromised hosts may acquire protection against MRSA infection on suppression of PMN-2 or increase in PMN-1 cells [69]. In the hematopoietic lineage, CMPs differentiates into immune cells along with myeloid-derived suppressor cells (MDSCs). MDSCs are heterogenous population of myeloid origin consisting of myeloid progenitors, immature granulocytes, immature monocytes and immature dendritic cells (Reviewed by Gabrilovich) [70]. Darcy et al. identified neutrophilic MDSCs characterised by the expression of CD66b high CD33 low CD11b high CD16 ± CD62L low HLA-DR low in circulation during sepsis pathogenesis. These MDSCs modulates T cell function and impairs T cell CD3 zeta-chain expression via L-arginine metabolism and contribute to the T cell dysfunction observed in sepsis [71]. In Mycobacterium tuberculosis infection, Manna et al. reported that the frequency of LDNs is associated with severity of tuberculosis. These LDNs displayed CD15 high CD33 high CD66b high CD16 low expression, produced higher level of ROS and phagocytic capacity was increased in comparison with autologous normal density neutrophils [72] [73]. Neutrophils in trauma patients showed distinct subsets characterized by CD62L low CD16 high levels that displayed hyper-segmented nuclei, indicating increased maturation (Fig. 4) [74]. CD62L low CD16 high neutrophils have also been observed in bronchoalveolar lavage fluid of infants with various types of viral respiratory infections [75]. CD49d + cysteinyl leukotriene receptor 1 (CysLTR1) + proatopic neutrophils which aid in the development of post-viral asthma were increased in nasal lavage in acute respiratory symptoms [76].
Neutrophil subsets referred as LDNs characterized based on buoyancy, have been extensively described in various infections (Fig. 2). Bowers et al. observed a significant increase in LDNs in HIV-1 infected subjects. These LDNs exhibited G-MDSC phenotype along with increased levels of PD-L1 in response to HIV-1 virions and participated in T-cell suppression via PD-L1/PD-1 signalling [17]. Human subjects infected with Severe Fever with Thrombocytopenia Syndrome Virus (SFTSV) displayed LDN subsets characterized by secreting pro-inflammatory cytokines, showed an ability to damage endothelial cells and which were responsible for phagocytosis of SFTSV [77].
Clinical studies in COVID-19 infected subjects showed that the neutrophil-to-lymphocyte ratio (NLR) correlated with disease severity and NLR was characterized by a reduced lymphocyte count and increased neutrophils [78][79][80][81]. Pre-clinical and clinical models have demonstrated two independent mechanisms for neutrophil activation during COVID-19 infections. SARS-CoV-2 infected alveolar epithelial cells release abundant levels of IL-6, IL-8, CXCL1 and CXCL2 resulting in the recruitment of neutrophils and in-turn, these activated neutrophils form extracellular traps and contribute to organ damage [82,83]. On the other hand, Veras et al. showed higher concentrations of NETs components in plasma, tracheal aspirates and lung autopsies in COVID-19 subjects. Mechanistically, authors demonstrated SARS-CoV-2 induced NETosis in healthy neutrophils in PAD-4 dependent manner [84]. Masso-silva et al. showed significantly higher levels of NETs in plasma and tracheal aspirate of patients hospitalized with COVID-19 and spontaneous NETs production was observed in SARS-CoV-2-infected lung airways and alveoli [85,86]. Neutrophil heterogeneity based on different densities, maturity and expression levels of surface markers have been observed in COVID-19 pathogenesis [28,87]. Carissimo et al. demonstrated a dramatic increase in immature neutrophils in COVID-19 patients which strongly correlated with the level of increased IL-6 and Hepatocyte growth factor (IP-10) which are crucial in driving cytokine storm and COVID-19 severity [88,89]. A strong correlation was observed between an increase in LDNs and disease severity in COVID-19 patients along with ability to produce NETs in comparison to healthy individuals [87,90]. LDNs expressing CD45 + CD66b + CD16 Int CD44 low CD11b Int were observed in severe COVID-19 patients with increased cytokine production, spontaneous NETs formation and enhanced phagocytic capacity. Increased immuno-suppressive CD16 bright /CD62L dim neutrophils in patients with pulmonary embolism on the day of ICU admission were

Systemic lupus erythematosus (SLE)
LDNs have also been attributed in SLE pathogenesis (Fig. 2) [92]. LDNs display pro-inflammatory phenotypes with increased production of IFN-g, TNF-α, and type I IFNs in SL and cause considerable endothelial cell cytotoxicity [94]. Increased production of type I IFNs by LDNs prevented the differentiation of endothelial progenitor cells to form mature endothelial cells [93]. In SLE subjects, LDNs were constitutively active, express elevated alarmins and immuno-stimulatory bactericidal proteins and were more prone to produce NETs leading to endothelial cell toxicity [12]. SLE patients showed elevated levels of LDNs, positively correlating with disease progression [94]. Noncalcified plaque burden (NCB) has been associated with LDNs in SLE patients and these activated LDNs contribute instability of coronary plaques [26].

Paradoxical effects of neutrophil sub-types in the pathogenesis of cancer
Neutrophils play an important and divisive role in the progression of solid tumors and the spread of malignancies. Neutrophil subpopulations have been identified with contrasting activities during tumor inflammation. Peripheral blood of cancer patients and experimental animals with tumors exhibit three subsets of neutrophils based on their density which includes high-density neutrophils (HDNs), LDNs and granulocytic-MDSCs. LDNs showed mature morphology and segmented nucleus whereas G-MDSCs display immature morphology with the expression of CD33 + /CD14 − /CD66b + /CD15 + /CD11b + / HLA-DR − / cell surface phenotype [95]. In subjects with advanced lung tumors, Shaul et al. using mass cytometry showed that 50% of LDNs expressed CXCR4 high CD10 low CD66 high PDL-1 high/inter indicating heterogeneity within LDN population [96]. Tumor associated neutrophils (TAN) showed hypersegmented nuclei with an increased ability to kill tumor cells and were characterized by elevated production of pro-inflammatory cytokines as shown in Fig. 3. TANs are differentiated into anti-tumor neutrophils (N1) and pro-tumor neutrophils (N2). N1 neutrophils exhibit anti-tumor properties. N2 neutrophils differentiated in TGF-β dependent manner show immature phenotype, increased arginase activity, inhibit T cell proliferation and promote tumor growth (Fig. 3) [97]. Zou et al. observed increased neutrophil counts in circulation, infiltration of TANs into a tumor and enhanced transition of TANs into N2 phenotype in vitro and in vivo models. IL-35, which is highly expressed in tumor tissue is responsible for the polarization of N2 phenotype and increased the neutrophil numbers in a tumor. IL-35 dependent IL-17 production served as a pro-tumorigenic factor which in turn elevated the expression of G-CSF and IL-6 leading to increased recruitment of neutrophils into the tumour microenvironment. IL-35 depleted the expression of TNFrelated apoptosis-inducing ligand (TRAIL) to increase the pro-angiogenic properties of neutrophils, and enhanced tumor growth and progression [98]. NK cells modulate the inflammatory property of neutrophils via IFN-γ-stimulated pathway to inhibit the expression of vascular endothelial growth factor-A (VEGF-A) which promoted angiogenesis and subsequent tumor growth in a rodent model [99]. Massena et al. identified pro-angiogenic subtype of neutrophils expressing VEGFR1 + , CxCR4 + , and CD49d + . Authors demonstrated that CD49d played a crucial role in the recruitment of neutrophils in response to VEGF-A and targeting this surface marker reduced the recruitment of proangiogenic neutrophils to hypoxic tissue [100]. Transcriptome analysis of three neutrophil subpopulations including naïve neutrophils, TANs and G-MDSCs revealed that TANs represented a distinct RNA profile [101]. TANs showed reduced expression of genes associated with oxidative burst while pro-inflammatory genes and antigen presenting complex genes were elevated [98]. In Gastric cancer patients, CD54 + TANs expressing high level of PD-L1, supress T cell activity via PD-1-PD-L-1 pathway and correlates with poor prognosis [102]. In hepatocellular carcinoma (HCC), CD66b + TANs exhibited increased expression of TNF-α, IL-8, CCL2 and cell-death ligand 1 (PDL1) and decreased CD62L expression. Prolonged survival and functional activity of TANs might be due to the higher expression of PDL1 through IL-6-STAT3-PDL1 signalling mechanism [103]. In melanoma models Huh et al. demonstrated IL-8 expressed by tumor cells increased the expression of β2 integrin and subsequently enhanced the interaction between neutrophils and melanoma cells. This interaction allowed the melanoma cells to transmigrate through the endothelium. IL-8 also helped in the retention of neutrophils in lung tissue [104]. These findings suggested that TANs are polarized from N1 phenotype to N2 phenotype depending on the tumor microenvironment indicting the plasticity and heterogeneity of neutrophils. TANs isolated from human lung tumors exhibited CD62L low CD54 hi phenotype with increased production in pro-inflammatory cytokines which enhanced T cell function and production of IFN-γ [105]. In colorectal tumor tissues, TANs exhibited similar morphology as normal neutrophils and displayed CD11b + CD33 + CD66b + CD45 + Lin − HLADR − phenotype with increased production of ROS and arginase 1 [106]. MDSCs observed in tumors of lung, bladder, head and neck exhibited low density phenotype, altered expression of cell surface markers, impaired effector functions and prolonged survival in comparison with normal neutrophils. These neutrophilic MDSCs lacking in chemokine receptors CXCR1 and CXCR2 which are essential for the extravasation of neutrophils led to reduced chemotaxis towards the tumour environment [107].

Neutrophil subtypes in autoimmune disorders
Mounting evidence implies that neutrophils play a crucial role in the course and severity of numerous autoimmune disorders [108]. Neutrophils in autoimmune disorders exhibit pro-inflammatory properties characterized by increased production of inflammatory mediators which subsequently lead to the production of autoantibodies and prime other leukocytes [109]. In SLE, apoptotic neutrophils are increased in correlation with pathological activity and anti-double-stranded DNA (anti-dsDNA) antibody levels [110]. Abnormal neutrophil adhesion and chemotaxis has been observed in rheumatoid arthritis [111]. Neutrophil heterogeneity and their functional relevance in various autoimmune disorders are discussed below.

Rheumatoid arthritis
Rheumatoid arthritis (RA) is a chronic autoimmune illness that causes inflammation in the joints and cartilage tissues. Peripheral blood neutrophils of RA subjects show distinct functionality from those in healthy individuals and are constitutively activated to produce ROS. Transcriptome analysis of RA neutrophils showed elevated expression of myeloblastin, TNF-α and membrane-bound receptor activator of nuclear factor κB (NF-κB) ligand (RANKL) [112]. Primed neutrophils from RA secreted cytokines such as B cell-activating factor (BAFF) and RANKL which activated B cells, osteoclasts and CD4 + T cells [113,114]. Cytokines including GM-CSF, IL-8 and TNF-α delayed neutrophil apoptosis and primed neutrophils to release granules in the synovial cavity [115]. RA neutrophils expressed high concentrations of elastase, gelatinase and collagenase responsible for tissue and cartilage damage [116]. Buckley et al., identified a phenotypically and functionally distinct subpopulation of neutrophils with the expression of CD54 high and CXCR1 low in RA individuals which showed a reverse transmigrated population [115]. These cells were different from tissue resident neutrophils which expressed CD54 low and CXCR1 low [115].

Multiple sclerosis
With an estimated prevalence of 2.5 million affected subjects globally, multiple sclerosis (MS) is the most common immune-mediated inflammatory illness affecting the central nervous system (CNS) [117]. Extravascular neutrophils expressing ICAM1 + in animal models of autoimmune encephalomyelitis (EAE) for multiple sclerosis were identified by Hawkins et al. (Fig. 4) [117]. These neutrophil subsets acquired macrophage like properties by showing MHC class II mediated antigen presentation and expressed aspartic peptidase retroviral-like 1 (ASPRV1, also known as SASPase) enzyme involved in autoimmune demyelination [118]. Neutrophils isolated from MS subjects expressed higher levels of CD43, IL-8R, n-formyl-methionyl-leucylphenylalanine and TLR-2 forming a distinct phenotype as shown in Fig. 4. These cells exhibited altered functionality with reduced apoptosis and oxidative burst, increased degranulation and NETs formation [119].

Systemic inflammatory response syndrome (SIRS)
Systemic inflammatory response syndrome (SIRS) is commonly observed in patients with severe burn injuries, pancreatitis, major surgery and polytrauma [69]. In addition to normal neutrophils (PMN), two other subsets PMN-I and PMN-II have been identified in SIRS patients. These cells differ in the expression of chemokine and cytokines. PMN-I displayed the expression of IL-12/CCL3 whereas PMN-II expressed IL-10/CCL2. PMN-I and PMN-II showed differential expression of toll-like receptors where PMN-I were characterized by expression of TLR8/TLR5/TLR4/TLR2, whereas PMN-II expressed TLR9/TLR7/TLR4/TLR2. These cells also differed in the distribution of cell surface markers, PMN-I displayed CD11b − CD49d + and PMN-II exhibited CD11b + CD49d − [69]. .

Current therapeutic approaches targeting neutrophils
Activated neutrophils are involved in many acute and chronic inflammatory diseases such as autoimmune disorders, cardiovascular diseases (thrombosis and atherosclerosis), respiratory diseases (COPD, asthma and ARDS), [120], neurological disorder (Alzheimer's and multiple sclerosis) [121,122], skin diseases (Behçet's disease and psoriasis) [123,124], metabolic diseases (obesity and diabetes mellitus) [125,126] and gastrointestinal diseases (inflammatory bowel and autoimmune hepatitis) [127]. Although, neutrophils possess a beneficial role in eliminating infections, over-functioning or failure to post-infection clearance of these cells causes significant tissue damage in aforementioned diseases. Hence, over the years, several studies have attempted to modulate over-functioning of neutrophils through pharmacological strategies. Wu et al. observed that CXCR2 antagonist SB225002 and theophylline induce a significant decrease in neutrophil viability and accelerated neutrophil apoptosis [128]. Food and Drug Administration (FDA) approved drug N-Acetylcysteine (Mucolytic drug) reduced ROS production in vitro and triggered a self-sustaining phlogogenic loop in the respiratory system [129]. Randomized, placebocontrolled, human study showed that AZD7986 (DPP1 inhibitor) inhibited whole blood neutrophil elastase activity [130]. Ali et al. reported that selective agonism of the adenosine A 2A receptor (CGS21680) suppresses antiphospholipid antibodies -mediated NETosis in protein kinase A-dependent fashion. CGS21680 also reduced thrombosis in the inferior venae cavae in in vivo models [131]. Lotamilast is a moderately potent PDE4 inhibitor (IC 50 = 2.8 nM) that effectively suppressed LPS induced neutrophilic pulmonary inflammation in mouse models [132]. Aikawa et al. demonstrated a clinical trial to treat ARDS by using inhibitor of neutrophil elastase Sivelestat (Elaspol, ONO 5046) [133].
NETs forming ability by neutrophils significantly varies during life time. Lipp et al. demonstrated that NETs forming ability was found reduced in infants (35-39 weeks) in comparison with healthy adults [134]. Neonates showed inability to form NETs because of expression of Neonatal NET-inhibitory factor (nNIF) that inhibited NETs formation by targeting the activity of peptidyl arginine deiminase 4 (PAD4), histone citrullination and nuclear condensation [135]. Further, nNIF administration was explored for inhibiting NETs in pathological states. nNIF improved the mortality in the CLP model of polymicrobial sepsis [135]. In murine model of ischemic stroke, nNIF delivery decreased ischemic stroke brain injury and related mortality [136]. Prostaglandins (PGE2) synthesized endogenously in arachidonic acid pathway inhibits PMA induced NETs formation through the activation of protein kinase A and cyclic AMP [137]. Tilgner et al. showcased the property of aspirin to reduce the production of NETs in C57BL/6 mice induced with acute lung injury (ALI) [138]. Li et al. showed the delay in the progression of multiple myeloma by inhibiting NETs formation upon targeting PAD4 using BMS-P5 [139]. In cystic fibrosis (CF), administration of rhDNase and airway clearance therapy are most frequently used techniques to mobilize sputum which contains large amount of extracellular DNA released by leukocyte specifically neutrophils. The rhDNase cleaves extracellular DNA, reduces viscoelasticity of sputum and mobilize sputum [140]. Authors involving 43 subjects with CF, demonstrated rhDNases improved the sputum mobilization [140]. A study involving 968 subjects with CF, demonstrated administration of rhDNases showed reduced respiratory exacerbations [141]. It is known that Gasdermin-D (GSDMD) is a key mediator of NETosis and a study consisting of 63 hospitalized patients with moderate and severe COVID-19 revealed higher expression of GSDMD genes [142]. In a mouse model of SARS-CoV-2 infection, the treatment with disulfiram (GSDMD inhibitor) inhibited NETs release and reduced organ damage [142].
In recent years, selective targeting of activated neutrophils in diseased models have been demonstrated. Employing nanotechnology Wang et al. delivered piceatannol, a small molecule that blocks β2 integrin pathway to prevent vascular inflammation through albumin nanoparticles which selectively targeted highly activated neutrophils attached to endothelium preventing neutrophil infiltration in murine models [143]. Bachmaier et al. identified two subsets of neutrophils based on endocytosis of albumin nanoparticles (ANP) (ANP high and ANP low ). ANP high neutrophils produced an inordinate amount of ROS and inflammatory cytokines. Authors targeted these ANP high neutrophils with ANPs loaded with piceatannol, a spleen tyrosine kinase (Syk) inhibitor to reduce the inflammation in sepsis and preserved neutrophilic host defense function in cecal ligation and puncture (CLP) mice model [144]. Activated neutrophils were also targeted utilizing α1-antitrypsin-derived peptide (surface decoration) to confer binding specificity to neutrophil elastase which enabled specific anchorage of nanoparticles to activated neutrophils [145]. In murine models of deep vein thrombosis, delivery of hydrochloroquine encapsulated in nanoparticles resulted in significantly smaller thrombi compared to either control or hydrocholoriquine alone [145].

Concluding remarks and future prospects
Genetic and epigenetic analysis revealed subsets of neutrophils expressing distinct transcriptional networks and concomitant differential gene expression and further displayed heterogeneous functions [24,61,62]. Existing data on neutrophil subtypes are based on transcriptome analysis, membrane markers and density. Neutrophils are terminally differentiated cells with decreased ability for transcription and also translation of nascent proteins. Studies have demonstrated transcriptional firing in response to stimuli such as LPS, PMA, A23187 and ionomycin [146]. However, proportions of these mRNA translated to proteins was unclear [146]. Recent studies exploring RNA sequencing has demonstrated existence of neutrophil subsets in both physiological and pathological states which may also indicate formation of neutrophil subsets in bone marrow itself and released to peripheral tissue with a pre-determined function [40,59,60,62]. Further, these subsets may acquire different states by post-translational modifications and metabolic states in response to patho(physiological) stimuli. Several layers of heterogeneity have been attributed to the formation of neutrophil subpopulations influenced by development, environment and activation state in both steady state and pathological conditions. Upstream pathogenic inducers and signalling mediators significantly vary among diseases and hence contribute to temporal diversity and dynamics of pathologically specific subsets and further attribute to the severity and duration of diseases. Hence, gene expression/ protein/post-translational modifications and metabolic intermediates also display disease specific profiles in activated neutrophils. Accordingly, insights into cellular and molecular mechanisms regulating functions of activated neutrophil subtypes/sub population may facilitate designing therapeutic targets specific to diseases. Moreover, inactive neutrophils are also gaining attention in several pathological conditions. Hence, designing anti-neutrophil therapies requires careful tailoring by only targeting subset of activated neutrophils and simultaneously maintaining their normal functioning. Hence, 'pan-anti-neutrophil' therapies and pharmacological blocking/activating specific signalling pathways might not be beneficial as pathological stimuli and associated signalling pathways may vary in neutrophil subsets during their activation in different diseases [33]. Inhibition of over functioning of neutrophils and simultaneously maintaining neutrophil homeostasis and restoring organ function may serve as a potential therapeutic strategy. For example, our earlier studies in the context of T2D show high glucose induces constitutive NETosis and leads to reduced response to infection [4]. Hence inhibition of glucose induced NETs formation and restoring the anti-microbial function of neutrophils to fight against infection may help in the clinical management of T2D associated infections. Generally, it is also experienced that neutrophils are activated while isolation which may be due to magnetic beads, density gradient compounds and mere centrifugation. Hence, full-fledged clinical studies using systems biology approaches involving the identification of epigenetic/genetic signatures along with transcriptomic, proteomic and metabolic profiles in neutrophil subsets in disease specific manner may help in a better understanding of neutrophil biology.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.