β2 integrin activation and signal transduction in leukocyte recruitment

Abstract

Leukocyte recruitment is a critical step in the pathogenesis of inflammatory and immunological responses. Cell adhesion molecules (CAMs) are involved in controlling cell movements and the recruitment process, and the integrin family of CAMs plays a key role. During cell movement, integrin function is dynamically and precisely regulated. However, this balance might be broken under pathological conditions. Thus, the functional regulation and molecular mechanisms of integrins related to diseases are often a focus of research. Integrin β2 is one of the most commonly expressed integrins in leukocytes that mediate leukocyte adhesion and migration, and it plays an important role in immune responses and inflammation. In this review, we focus on specific functions of integrin β2 in leukocyte recruitment, the conformational changes and signal transduction of integrin β2 activation, the similarities between murine and human factors, and how new insights into these processes can inform future therapies for inflammation and immune diseases.

INTRODUCTION

Blood leukocytes, which include neutrophils, monocytes, lymphocytes, and others, play critical roles in immune defense to pathogens and inflammatory responses. Blood leukocytes are recruited to tissues during inflammation and infections in a step-by-step process called the leukocyte recruitment cascade (1, 2) (Fig. 1). This includes rolling, slow-rolling, arrest, spreading, intravascular crawling, trans-endothelial migration, and migration to the site of inflammation. Integrins are αβ-heterodimer adhesion molecules expressed on leukocyte surfaces (Fig. 2A). β2 integrins, including lymphocyte function-associated antigen 1 (LFA-1, also known as α_Lβ₂ and CD11a/CD18), macrophage-1 antigen (Mac-1, also known as α_Mβ₂, CD11b/CD18, and complement receptor 3), α_Xβ₂ (also known as CD11c/CD18 and complement receptor 4), and α_Dβ₂ (also known as CD11d/CD18), are leukocyte-specific integrins (3, 4) and are involved in most steps of the leukocyte recruitment cascade, which will be introduced in detail in ROLE OF β2 INTEGRIN IN LEUKOCYTE RECRUITMENT. β2 integrin deficiencies in humans lead to leukocyte adhesion deficiency (LAD) syndromes (5). Mutations in β2 integrins cause LAD-I, which presents recurrent, life-threatening bacterial infections. Mutations in a critical integrin activation adaptor protein kindlin-3 result in LAD-III, characterized by severe bacterial infections and a severe bleeding disorder.

Figure 1.

Schematic showing the involvement of β2 integrins in the leukocyte recruitment cascade. Flowing leukocytes roll on the vascular endothelial surface. β2 integrins contribute to slow-rolling and firm adhesion (arrest). Leukocytes then spread on endothelial cells, perform intravascular crawling, and transmigrate through endothelial cells (paracellular or transcellular). Leukocytes may swarm during in-tissue migration to the site of inflammation or infection. LFA-1, lymphocyte function-associated antigen 1; Mac-1, macrophage-1 antigen.

Figure 2.

Integrin conformational changes and signaling pathways during activation. A: schematic showing allosteric pathways of β2 integrin activation. B: schematic showing G protein-coupled receptor (GPCR) and P-selectin glycoprotein ligand-1 (PSGL-1)-triggered β2 integrin activation signaling pathways. CalDAG-GEFI, calcium and DAG-regulated guanine nucleotide exchange factor I; CT, cytoplasmic tail; DAG, diacylglycerol; E⁻H⁻, bent low-affinity; E⁺H⁻, extended low-affinity; E⁻H⁺, bent high-affinity; E⁺H⁺, extended high-affinity; Gα, G protein α subunit; Gβγ, G protein βγ complex; ICAMs, intercellular adhesion molecules; MAPK, mitogen-activated protein kinase; PIP1, phosphatidylinositol 4-phosphate; PIP2, phosphatidylinositol 4,5-bisphosphate; PIP3, phosphatidylinositol (3–5)-trisphosphate; PIP5K1C, phosphatidylinositol-4-phosphate 5-kinase type-1-γ; PI3K, phosphoinositide 3-kinase; Rap1, Ras-proximate-1 or Ras-related protein 1; Rap1-GEF, Rap1 guanine nucleotide exchange factor; RIAM, Rap1-GTP-interacting adapter molecule; SLP-76, SH2 domain-containing leukocyte protein of 76 kDa; TM, transmembrane.

β2 INTEGRIN ACTIVATION AND CONFORMATIONAL CHANGES

Both integrin α and β subunits include a large ectodomain consisting of a headpiece and a tailpiece, a transmembrane domain, and a short flexible cytoplasmic tail (6) (Fig. 2A). The ligand-binding site of β2 integrins is on the αA/I domain. The conformational changes of β2 integrins are strictly regulated. When cells are not stimulated, β2 integrins are in a bent low-affinity conformation (E⁻H⁻), and the binding of E⁻H⁻ β2 integrins and ligands is very weak. The “switchblade” model is a widely accepted model for integrin activation (6), and the activation process includes two steps: 1) the extension of integrin ectodomain (E⁺) and 2) the rearrangement of ligand-binding sites that acquire high affinity to ligands (H⁺). After activation, the β2 integrins present an extended high-affinity conformation (E⁺H⁺) (6). The conformation between the above two conformations is intermediate affinity, which has an extended low-affinity headpiece conformation (E⁺H⁻) (6, 7). The switchblade model suggests a sequence of conformational changes from E⁻H⁻ to E⁺H⁻ to E⁺H⁺ (6). During leukocyte migration, the transition from E⁺H⁻ to E⁺H⁺ might involve intracellular forces applied on the integrin β-subunits through bond cytoskeletons (8).

Another integrin activation model is the “deadbolt” model (9), which involves a bent high-affinity conformation (E⁻H⁺) of integrin αVβ3 in a binding complex with fibronectin (10). In this model, a hairpin loop in the β tail domain acts as a deadbolt to restrain the displacement of the β-A/I domain β6-α7 loop and maintain integrin in the H⁻ state. The movement of this loop results in E⁻H⁺ conformation and allows ligand binding. Then, the binding of ligands provides a pulling force that extends the ectodomain. This model is controversial in that the “deadbolt” works on Mac-1 (11) but not β3 integrins (12). However, E⁻H⁺ α_Xβ₂ has been reported (13). Recent studies using quantitative dynamic foot printing (qDF) microscopy and stochastic optical reconstruction microscopy showed that primary human neutrophils express E⁻H⁺ β₂ integrins that bind intercellular adhesion molecules (ICAMs) in cis and auto-inhibit integrin activation and neutrophil adhesion (7, 14). With qDF imaging, the conformational changes from E⁻H⁻ to E⁺H⁻ and then to E⁺H⁺ that favor the switchblade model and the conformational changes from E⁻H⁻ to E⁻H⁺ and then to E⁺H⁺ that favor the deadbolt model were both observed. Another study showed that Mac-1 could bind Fc-γ receptor IIA (FcγRIIA, CD32) in cis and limit antibody-mediated neutrophil recruitment (15), further demonstrating that E⁻H⁺ Mac-1 exists. The mechanism of how this E⁻H⁺ activation occurs requires further investigation.

ROLE OF β2 INTEGRIN IN LEUKOCYTE RECRUITMENT

Leukocyte recruitment from blood (Fig. 1 and Table 1) is initiated by rolling on the vascular endothelium mediated by the interactions of leukocyte P-selectin glycoprotein ligand-1 (PSGL-1) and endothelial selectins (1, 2). β2 integrins are required to slow the rolling of cells, a process called “slow-rolling.” Slow-rolling is mediated by E⁺ integrins triggered by the interaction of PSGL-1 and selectins, a signaling pathway that will be introduced in a later section. Slow-rolling depends on LFA-1 but not Mac-1 interacting with intercellular adhesion molecule 1 (ICAM-1) in mouse and human neutrophils (Table 1).

Table 1.

Molecules involved in neutrophil recruitment cascades in mice and humans

Recruitment Cascades	Human or Mouse	Involved β2 Integrins	Involved Ligands	Involved Integrin Adaptor Proteins	Integrin Activation State Required	References
Slow-rolling	Mouse	LFA-1	ICAM-1	Talin-1	ND	(16–19)
	Human	LFA-1	ICAM-1	Talin-1	E+	(20)
Arrest	Mouse	LFA-1	ICAM-1	Talin-1, kindlin-3	ND	(16, 17, 21–23)
	Human	LFA-1, Mac-1	ICAM-1	Talin-1, kindlin-3	E+H+	(7, 24)
Spreading	Mouse	Mac-1	ICAM-1, immune complex	Kindlin-3	ND	(22, 25, 26)
	Human	ND	ICAM-1	Talin-1, kindlin-3	E+H+	(27, 28)
Intravascular crawling	Mouse	Mac-1	ICAM-1	NA	ND	(29)
Trans-endothelial migration	Mouse	LFA-1, Mac-1	JAM-A, JAM-C, ICAM-1	NA	ND	(30–33)
	Human	LFA-1	JAM-A	NA	ND	(34)
Migration /chemotaxis	Human	LFA-1, Mac-1	ICAM-1, collagen, fibronectin, β-glucan	Talin-1, kindlin-3	ND	(35–38)
Swarming	Mouse	LFA-1, Mac-1	ND	Talin-1	ND	(39)

ICAM-1, intercellular adhesion molecule 1; JAM-A, junctional adhesion molecule-A; JAM-C, junctional adhesion molecule-C; LFA-1, lymphocyte function-associated antigen 1; Mac-1, macrophage-1 antigen; NA, not applicable; ND, not determined.

When rolling leukocytes encounter chemokines expressed on endothelial surfaces or in blood, the G protein-coupled receptor (GPCR)-mediated signaling pathway (which will be discussed more in CHEMOKINE-GPCR-INDUCED SIGNALING PATHWAYS IN β2 INTEGRIN ACTIVATION) induces fully E⁺H⁺ β2 integrins that bind ligand in trans and stops rolling cells, a process called “arrest” (Fig. 1). LFA-1 but not Mac-1 interacting with ICAM-1 is important for mouse neutrophil arrest, and both LFA-1 and Mac-1 interacting with ICAM-1 are important for human neutrophil arrest (Table 1).

After arrest, round leukocytes undergo large morphological changes and spread on the endothelium surface, which requires β2 integrins as well (27). Then leukocytes crawl in the vascular lumen. Neutrophils use Mac-1 to perform intravascular crawling (Table 1). Both LFA-1 and Mac-1 are important for monocyte intravascular crawling (40).

Leukocyte trans-endothelial migration is the next step of recruitment. Trans-endothelial migration may happen at the junctions between endothelial cells (paracellular) or in the middle of an endothelial cell (transcellular) (30). A study using LFA-1-deficient mice and blocking antibodies in an in vitro system (cultured endothelial cell monolayers) showed that the trans-endothelial migration of neutrophils and T lymphocytes are LFA-1-dependent, whereas monocytes use both LFA-1 and Mac-1 for trans-endothelial migration (Table 1). However, later studies showed that Mac-1 is also important for neutrophil trans-endothelial migration by interacting with endothelial junctional adhesion molecule C (JAM-C or JAM-3, Table 1). Endothelial JAM-A (or JAM-1) is the ligand of LFA-1 involved in trans-endothelial migration of neutrophils and T lymphocytes (Table 1).

Leukocyte migration within tissues is a mechanistically different process than in the bloodstream that can occur without β2 integrins (41, 42). However, β2 integrins also play critical roles during in-tissue migration of leukocytes in some cases. Antibody blockade of Mac-1 can decrease neutrophil migration through low-concentration (0.4 mg/mL) but not high-concentration (>0.6 mg/mL) collagen gels (Table 1). Leukocytes migrate to the site of infection or sterile injury and form swarms, a process called “swarming.” Swarming is visualized by intravital microscopy of the dermis (Table 1). Using integrin β2, αL, αM, and talin-1 (a key adaptor of integrin activation; details will be discussed in INTEGRIN BINDING ADAPTOR PROTEINS) knockout mice, it was shown that both LFA-1 and Mac-1 are involved, and integrin activation is critical for neutrophil swarming (Table 1). It has also been shown that inhibition of LFA-1/ICAM-1 reduces the speed of interstitial T cell migration within ex vivo lymph nodes (43). Intravital imaging of lymph nodes showed that LFA-1 contributes to subtle cell elongation during native T cell migration (44). A T cell migration study using emission anisotropy total internal reflection fluorescence microscopy showed that cytoskeleton and ligand-bound integrins orient in the same direction as retrograde actin flow with their cytoskeleton-binding β-subunits tilted by the applied force, suggesting a molecular model of integrin activation (E⁺H⁻ to E⁺H⁺) by cytoskeletal force (8).

CHEMOKINE-GPCR-INDUCED SIGNALING PATHWAYS IN β2 INTEGRIN ACTIVATION

Chemokines can regulate integrin-mediated cell adhesion and migration through the inside-out signaling pathway (Fig. 2B). This signaling is generally induced by chemokine stimulation through corresponding GPCRs, such as chemokine C-X-C motif ligand 8 [CXCL8, also known as interleukin-8 (IL-8)] and neutrophil C-X-C motif chemokine receptor 2 (CXCR2) (45), chemokine C-C motif chemokine ligand 2 [CCL2, also known as monocyte chemoattractant protein-1 (MCP-1)] and monocyte C-C motif chemokine receptor 2 (CCR2) (46), and CXCL12 [also known as stromal cell-derived factor 1 (SDF-1)] and lymphocyte CXCR4 (47). The activation of GPCRs dissociates G protein to Gα and Gβγ subunits (48). Gα activates downstream Ras homolog gene family (Rho) GTPases, such as Ras homolog gene family, member A (RhoA) and cell division control protein 42 (CDC42), and leads to β2 integrin activation (49). Blockade of the 23–40 RhoA effector region inhibits chemokine stimulation-induced high-affinity integrin αLβ2, reduces the adhesion of αLβ2 to high concentrations of ICAM-1, and blocks leukocyte homing to Peyer’s patches in vivo (50). Gβγ activates another Rho GTPase downstream, Ras-related C3 botulinum toxin substrate 1 (Rac1), then activates phospholipase C β2 (PLCβ2) and PLCβ3, induces intracellular Ca²⁺ flux, and activates β2 integrin (51). Other Rho GTPases, such as Rac2 (52), and phospholipases, such as PLCγ (53) and phospholipase D1 (PLD1) (4), may be involved in the integrin activation signaling pathway. Downstream of Rho GTPases, phospholipases, and intracellular Ca²⁺ flux in the integrin activation signaling pathway are Ras-related protein 1 (Rap1) GTPases (53–55), which are required for leukocyte adhesion (55). The activation of Rap1 by PLC through Ca²⁺ and diacylglycerol (DAG) depends on Rap1 guanine nucleotide exchange factors (Rap1-GEFs) (55) and calcium and DAG-regulated guanine nucleotide exchange factors (CalDAG-GEFs) (54). Another mediator between PLC and Rap1 may be p38 mitogen-activated protein kinase (MAPK) (56). N-formyl-Met-Leu-Phe stimulation activates neutrophil p38 MAPK and depends on PLC (57). Phosphatidylinositol-4-phosphate 5-kinase type I-γ (PIP5K1C) is downstream of PLD1 (21).

Rap1-GTP-interacting adaptor molecule (RIAM) is a Rap1 binding protein important for leukocyte β2 integrin activation. Loss of RIAM in mice results in impaired β2 integrin activation and leukocyte adhesion (58). RIAM binds the Src kinase-associated phosphoprotein 2 (Skap2), which is activated by phosphatidylinositol (3–5)-trisphosphate (PIP3) (59), and the adhesion and degranulation-promoting adapter protein (ADAP) module to promote membrane targeting of the RIAM-Rap1 complex, which is required for β2 integrin activation of T cells (60). RIAM binds talin-1, which is an integrin-binding activation adaptor (discussed in detail below) and recruits talin-1 to the membrane to bind and activate integrins (61). Lamellipodin (Lpd) is another bridge between Rap1 and talin-1 (61). RIAM and Lpd distinctly regulate integrin activation on effector and regulatory T cells, respectively (62).

Skap2, which is a RIAM-binding protein mentioned in the above paragraph, is important for regulating β2 integrin activation in neutrophils (63). Besides RIAM and ADAP, Skap2 was also found to bind to Wiskott–Aldrich syndrome protein (WASp), mediate actin polymerization, and be required for the recruitment of talin-1 and kindlin-3 to the β₂ integrin cytoplasmic domain (63).

INTEGRIN BINDING ADAPTOR PROTEINS

Talin-1 is a key adaptor protein for integrin activation. It consists of a 50 kDa N-terminal head domain (THD) and a 220 kDa rod domain (64). The THD of talin-1 is a FERM domain (F for 4.1 protein, E for ezrin, R for radixin, and M for moesin) consisting of four subdomains, F0–F3. The F3 subdomain contains two binding sites (identified by three mutations, L325R, R358A, and W359A): one that binds to the membrane-proximal α-helix and one that binds to the membrane-proximal NPxY motif of the integrin β cytoplasmic tail (64). The W359A mutation eliminates the binding of talin-1 to β2 integrin and shows complete deficiencies of leukocyte slow-rolling and arrest (17), which is similarly observed in talin-1 knockout leukocytes (16). The talin-1 mutant L325R still binds to β2 integrin but shows a deficiency in leukocyte arrest but not slow-rolling (17), suggesting that talin-1 binding to the membrane-proximal α-helix of integrin β2 is required for H⁺ but not E⁺. The binding sites on the THD are hidden by the interaction with the rod domain of inactive talin-1 (65). Several molecules have been reported to disrupt the interactions between the THD and the rod domain, such as PIP5K1C-generated phosphatidylinositol 4,5-bisphosphate (PIP2, which is a cell membrane component) (66), calpain (67), and RIAM (68). Interaction of talin with the cell membrane is also essential for integrin activation. Mutation of K322 (K322D), which is predicted to be directed toward the membrane, but not K320 (K320D), which is predicted to point away from the membrane, on the talin-1 F3 domain abolished integrin activation (65). Mutations of the positively charged patch in the F2 domain of talin-1 (K256E/K272E/K274E/R277E) abolished the binding to phosphatidylcholine in vitro and αIIbβ3 integrin activation in Chinese hamster ovary (CHO) cells (69), suggesting the recruitment of talin-1 to cell membrane phosphatidylcholine is important for integrin activation. Another study showed that point mutations of the basic patch in the talin-1 F2 domain (K272A/K274Q/R277E) or pair-wise mutations of the lysine residues of the basic finger in the talin-1 F3 domain (K320A/K322A; K322A/K324A) reduced the association of talin-1 head domain with the PIP2-containing lipid surface up to sixfold, as measured by the response units at equilibrium (70). This study also showed that the talin-1 F2 domain K272A/K274Q/R277E mutation and F3 domain K322A/K324A mutation but not the F3 domain K320A/K322A mutation blocked integrin clustering (70), suggesting the role of PIP2 and talin-1 in integrin clustering. The lipid-binding-deficient talin-2 mutation (K272Q/K316Q/K324Q/E342Q/K343Q) failed to bind PIP2 in vitro (71). Experiments in CHO cells demonstrated that this mutant failed to be recruited to the cell membrane and activate αIIbβ3 integrins (71). However, all these studies were done on integrin αIIbβ3. The role of talin-PIP2 interaction needs to be further explored on leukocyte β2 integrins. Talin-1 is required for both β2 integrin E⁺ and H⁺, which are critical for leukocyte slow-rolling and arrest, respectively (16).

Kindlin-3 is another adaptor protein that binds to the NPxY motif of the β2 integrin cytoplasmic tail (72) and regulates β2 integrin activation in leukocytes (22). Similar to talin-1, kindlin-3 also contains a FERM domain that binds to the distal NPxY motif, the β cytoplasmic tail, and can be further divided into four subdomains, F0–F3. Kindlin-3-deficient mouse neutrophils showed less recruitment in a phorbol ester-treated ear inflammation model, similar rolling velocity but less adhesion on postcapillary venules of cremaster muscles, and less adhesion and spreading on an ICAM-1-coated surface in vitro (22). In another study (16), kindlin-3-deficient mouse neutrophils showed less binding of soluble ICAM-1 after CXCL1 stimulation. Consistent with the previous study, kindlin-3-deficient mouse neutrophils showed similar rolling velocity on the substrate of both E-selectin and E-selectin plus ICAM-1 compared with wild-type controls. Similar to wild-type neutrophils, kindlin-3-deficient neutrophils showed slow-rolling by rolling slower on the substrate of P-selectin plus ICAM-1 compared with that of P-selectin only, indicating kindlin-3 is dispensable for β2 integrin E⁺. This was also confirmed by intravital cremaster muscle imaging showing that LFA-1 blockade increased the rolling velocity of both wild-type and kindlin-3 knockout neutrophils. Kindlin-3 is required for neutrophil arrest in vitro and in vivo, indicating that kindlin-3 is necessary for β2 integrin H⁺. Kindlin-3 knockdown in HL60 cells inhibited expression of H⁺ but not E⁺ β2 integrins after stimulation with constitutively active Rap1a peptide. A Q597A/W598A mutation in the F3 domain of mouse kindlin-3 (equivalent to human kindlin-3 Q599A/W600A) disables the interaction of kindlin-3 and β cytoplasmic tail and suppresses neutrophil adhesion (73). The pleckstrin homology (PH) domain is required for the recruitment of kindlin-3 to the cell membrane, which is essential for β2 integrin activation and adhesion of B cells (74) and neutrophils (24). The PH domain itself is sufficient to recruit kindlin-3 to the cell membrane in response to chemokine stimulation (24). Dynamic live-cell imaging showed that kindlin-3 is recruited to the cell membrane before H⁺ of β2 integrins (24). Transfecting kindlin-3 back into kindlin-3 knockout cells restores neutrophil adhesion in vivo (24). Kindlin-3 PH domain shows a high affinity to PIP3 but not PIP2 (75). PIP2 is phosphorylated at the 3' site by phosphoinositide 3-kinases (PI3Ks) to produce PIP3 (76). Chemokine binding to GPCRs activates PI3K through the release of Gβγ (76, 77). PI3K has been linked to neutrophil migration (77), and PI3Kγ knockout neutrophils showed a significant decrease of adhesion in vivo (78). Thus, PI3Kγ might be the upstream factor that regulates kindlin-3 membrane recruitment. It has been shown that pretreating cells with PI3K inhibitors wortmannin and LY294002 significantly inhibits chemokine-induced leukocyte adhesion on ICAM-1. Further tests showed that PI3K inhibition results in less integrin clustering but similar soluble ICAM-1 binding in response to chemokine stimulation, suggesting that PI3K is involved in regulating β2 integrin mobility/clustering/avidity but not affinity changes (79).

PSGL-1-SELECTINS INDUCE SIGNALING PATHWAYS IN β2 INTEGRIN ACTIVATION

PSGL-1, also known as CD162, is an adhesion molecule expressed on leukocytes (1, 2). PSGL-1 molecules interact with selectins, including P-selectin (CD62P), E-selectin (CD62E), and L-selectin (CD62L), and mediate leukocyte rolling on endothelial cells. It has been found that PSGL-1-selectin interaction can induce E⁺ of β2 integrins (Fig. 2B) (7, 20, 80). These intermediate-affinity E⁺H⁻ β2 integrins interact with ICAM-1 and slow down leukocyte rolling (7, 20). This activation pathway involves activation of Src family kinases (SFKs), DNAX-activating protein of 12 kDa (DAP12), FcγR, spleen tyrosine kinase (Syk), and p38 MAPK (80). Src homology 2 domain-containing leukocyte phosphoprotein of 76 kD (SLP-76) and adhesion and degranulation promoting adaptor protein (ADAP) are involved in PSGL-1-E-selectin-mediated integrin activation (81). SLP-76 is required for the activation of Bruton’s tyrosine kinase (Btk), PLCγ2, PI3Kγ, and p38 MAPK (81), and is downstream of SFKs (82). Tyrosine 145, rather than tyrosines 112 and 128, on SLP-76 is phosphorylated in the PSGL-1-selectin-induced signaling pathway (82). Rap1a activation through CalDAG-EGFI and p38 MAPK has been shown in E-selectin-dependent integrin activation (56). L-selectin expressed on leukocytes can engage with PSGL-1 and favor PSGL-1-induced integrin activation (83). Similar to PSGL-1, L-selectins that directly bind to E-selectins during leukocyte rolling can trigger E⁺ of β2 integrins (84). Clustered L-selectins that bind with E-selectins can induce H⁺ of β2 integrins (84). As talin-1 is involved in PSGL-1/selectin-mediated slow-rolling (16), talin-1 recruitment to the β2 integrin cytoplasmic tail may happen in PSGL-1-selectin-induced signaling pathways.

CONCLUDING REMARKS

As β2 integrins are critical for leukocyte recruitment, drugs targeting β2 integrins have been developed to treat inflammatory diseases (85). However, pan-β2 integrin targeting results in severe side effects. Efalizumab, an LFA-1 monoclonal antibody to treat autoimmune diseases, has side effects including bacterial sepsis, viral meningitis, invasive fungal disease, and progressive multifocal leukoencephalopathy, and was withdrawn from the market in 2009 (85). Thus, drugs targeting specific integrin conformations or signaling regulators may be safer and effective.

Although many signaling regulators have been discovered, as we discussed above, identifying new signaling regulators as potential drug targets remains important. For example, E⁻H⁺ β2 integrins show an auto-inhibitory function in integrin activation and leukocyte recruitment (7, 15), but the signaling regulators for E^–H⁺ β2 integrins are not known. Omics studies of β2 integrin activation loss-of-function and gain-of-function systems may be helpful. Monocytes from patients with cystic fibrosis showed a deficiency in β2 integrin activation (86). Thus, comparing the transcriptomes (87, 88) or proteomes (89) of cystic fibrosis and healthy monocytes might be a way to identify new signaling regulators of integrin activation. As a model of neutrophils, HL60 cells need to be differentiated to activate their β2 integrins. Therefore, transcriptomic (90, 91) or proteomic studies of HL60 may also provide further insight.

GRANTS

This research was supported by funding from the National Institutes of Health (NIH, R01HL145454) and a startup fund from UConn Health.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Z.F. prepared figures; H.S. drafted manuscript; H.S., L.H., and Z.F. edited and revised manuscript; Z.F. approved final version of manuscript.