Neutrophil Signaling That Challenges Dogmata of G Protein-Coupled Receptor Regulated Functions

Abstract

Activation as well as recruitment of neutrophils, the most abundant leukocyte in human blood, to sites of infection/inflammation largely rely on surface-exposed chemoattractant receptors. These receptors belong to the family of 7-transmembrane domain receptors also known as G protein-coupled receptors (GPCRs) due to the fact that part of the downstream signaling relies on an activation of heterotrimeric G proteins. The neutrophil GPCRs share significant sequence homologies but bind many structurally diverse activating (agonistic) and inhibiting (antagonistic) ligands, ranging from fatty acids to purines, peptides, and lipopeptides. Recent structural and functional studies of neutrophil receptors have generated important information on GPCR biology in general; this knowledge aids in the overall understanding of general pharmacological principles, governing regulation of neutrophil function and inflammatory processes, including novel leukocyte receptor activities related to ligand recognition, biased/functional selective signaling, allosteric modulation, desensitization mechanisms and reactivation, and communication (cross-talk) between GPCRs. This review summarizes the recent discoveries and pharmacological hallmarks with focus on neutrophil GPCRs. In addition, unmet challenges are dealt with, including recognition by the receptors of diverse ligands and how biased signaling mediates different biological effects.

1. General Features and Recent Advances Regarding Ligand Recognition and Downstream Signaling of G Protein-Coupled Receptors (GPCRs)

1.1. Basic Characteristics of GPCRs

Membrane receptors expressed at the plasma membrane of cells interact with their specific signaling ligands present in their environment, and these interactions constitute the basis for how cells recognize and react to a large variety of such molecules. The receptor-binding compounds are commonly not membrane-permeable and include hormones, growth factors, and neurotransmitters. With respect to the receptors expressed in neutrophil phagocytes, these are designed to recognize a large number of mediators that regulate basic cell functions as well as inflammatory reactions. The family of 7-transmembrane receptors (7TMR), or G protein-coupled receptors (GPCRs), represent the largest family of receptors exposed on cell surfaces.¹ The peptide chain of a plasma membrane localized GPCR spans the membrane seven times; by that means, the receptor exposes extracellular domains that are designed to recognize agonistic/antagonistic ligands and cytoplasmic domains that transfer cell activating/inhibiting signals of importance for many biological processes including immune reactivity (Figure 1 and refs (2−4)). The fact that the GPCRs are integrated into the phospholipid layer in cell-membranes makes it hard to obtain structural data with high resolution for this receptor group. Despite this, over 800 sequence-based comparative structural predictions exist to date. In addition, based on work rewarded with a Nobel Prize in 2012, a substantial number of crystal structures for around 30 receptors with and without bound ligand and/or interacting signaling proteins are also available.^5,6

Figure 1.

Schematic description of granule- and membrane-localized 7-transmembrane GPCRs, membrane proteins having seven membrane-spanning α-helices. On the cell surface, the N-terminus and three loops are exposed, and these parts are connected with three intracellular loops and the C-terminal tail facing the cytosol. Conventional agonist/antagonists bind to the surface-exposed orthostetic binding site, and signaling is initiated by the heterotrimeric G protein present on the cytosolic side of the cell membrane. In resting/naïve peripheral blood neutrophils (shown to the left), most of the formyl peptide receptors (FPRs) are stored together with many other receptors and adhesion molecules in secretory vesicles of endocytic origin and in specific granules formed during differentiation in the bone marrow. These intracellularly stored receptors are mobilized to the cell surface during tissue recruitment and/or activation of the cells.

The rhodopsin family is by far the largest GPCR subgroup and includes a large number of receptors related to sense of smell (olfactory). In addition, the rhodopsin family also includes a number of well-characterized nonolfactory GPCRs, among which the adrenergic receptors, opioid receptors, and dopamine receptors can be mentioned. The transmembrane parts of GPCRs are α-helices, and these are connected by three loops exposed on cell surface and three loops facing the cytosol. The N-terminal tail of plasma-membrane-localized receptors is exposed extracellularly, whereas the C-terminal tail is cytoplasmic (Figure 1). To transfer the cellular responses from receptor specific ligands that bind to the extracellular/transmembrane localized binding cavity of the receptor, GPCRs require intracellular coupling to a heterotrimeric G protein containing two different parts termed the α- and β/γ-subunit, respectively.⁷ The receptors that regulate immune-cell-function include formyl peptide receptors (FPRs), free fatty acid receptors (FFARs), and the different chemokine receptors which all belong to the rhodopsin family of GPCRs. The specific ligands recognized by these receptors are numerous and diverse, ranging from cytokines and small molecules to lipids and peptide chains of different sizes (peptides/polypeptides/proteins).²⁻⁴ It should also be noted that in innate immune cells such as the neutrophil granulocytes many receptors are stored in intracellular/secretory organelles in which the N-terminal tail faces the vesicle matrix compartment (Figure 1). These receptors are first available for ligand binding after an induction of secretion, a process that results in a mobilization of the receptors from the storage organelles to the cell surface.

1.2. Primary Downstream Signaling of GPCRs Activated by Orthosteric Agonists

Ligands that bind to specific sites present on the surface-exposed parts of GPCRs, normally interact with a designated orthosteric binding pocket, typically formed by the extracellular tail/loops and parts of the transmembrane domains reachable from outside the cell.⁸ When the interacting ligand is an agonist, a change in receptor structure/conformation is induced. The structural change initially involves the cytosolic parts of the receptor in contact with a heterotrimeric G protein present on the cytosolic side of the membrane. The Gα and Gβγ subunit parts constituting the trimeric G protein are separated from each other upon activation, and these subunits initiate distinct as well as overlapping downstream signals; the Gα parts of the different G proteins have an intrinsic GTPase activity, and this keeps the signaling capacity of receptors at a very low or zero level in the absence of a bound agonist. When linked to a Gα subunit, the β/γ subunits form a single inactive complex. Binding of an agonist to a resting/nonsignaling receptor induces conformational changes in the receptor that promote a downstream exchange of guanosine diphosphate (GDP) bound to the G protein, for guanosine triphosphate (GTP), and this then leads to a dissociation of the G protein subunits and an activation of Gα. When the Gβγ subunit is bound to a Gα subunit, an activation of Gα signaling is prevented, but once in an unbound form, the Gβγ subunit is also involved in signaling. Several proteins, both bound to the membrane and free in the cytoplasm, can interact with the separated G protein subunits and regulate their activities as well as further transduce signaling.⁹

1.3. Signaling Cascade of Activated GPCRs

The receptor downstream signaling cascades triggered by different subtypes of G proteins overlap, and the responses at the cellular level triggered by activated GPCRs depend not only on the precise receptor but also largely on the cell type in which the receptor is expressed. In addition, factors like the particular ligand used, the concentration of the ligand, costimulatory influences, and the state of the responding cell are of importance for signaling and the functional repertoire. A general and very common signal generated by various receptors in many cells, is the release of calcium from intracellular storage organelles leading to increased levels of free calcium ions in the cytoplasm ([Ca²⁺]_i). A transient rise in [Ca²⁺]_i is one of the earliest events coupled to activation of GPCRs (Figure 2). A phospholipase (PLCβ) located immediately downstream of the activated receptor initiates this signaling pathway. The activated PLC hydrolyzes PIP₂ (phosphatidylinositol bisphosphate) into inositol trisphosphate (IP₃) and diacylglycerol (DAG). When IP₃ binds its receptor exposed on the Ca²⁺-storing organelle (in many cells the endoplasmic reticulum (ER)), a transient intracellular calcium release from intracellular stores is promoted. In addition, the increased level of Ca²⁺ participates, together with DAG formed during hydrolysis of PIP2, in the activation of protein kinase C (PKC). The storage organelles also initiate an influx of Ca²⁺ through the plasma membrane; not the increase in the cytosolic concentration of Ca²⁺, but the depletion of Ca²⁺ in the storing organelle is sensed by the plasma membrane and triggers an opening of store-operated Ca²⁺ channels (SOCs) in that membrane.¹⁰ This second increase in [Ca²⁺]_i further activates other downstream signaling systems.¹¹ The release of Ca²⁺ from storage organelles into the cytosol has been regarded as a signaling hallmark of receptors that use heterotrimeric G proteins containing a Gαq subunit. It should, however, be noticed that the same signaling pathway may also be triggered by Gβγ subunits of Gαi-linked GPCRs.^12,13 There are four main G protein subtypes (Gαq, Gαi/Go, Gαs, and G_12/13) of which Gαi and Gαq are regarded as the most important for neutrophil function; there are also a variety of isoforms among the Gβγ subunits as well as different proteins that regulate their functions.^14,15

Figure 2.

Schematic description of the early Ca²⁺ signaling pathway initiated by activated GPCRs. Activation of phospholipase (PLCβ) immediately downstream of an agonist-occupied GPCR is mediated either by the Gαq-subunit of the activated G protein (signaling to the left) or by the Gβγ subunit of Gαi-linked GPCRs (signaling to the right). PLC hydrolyzes PIP₂ to generate DAG and IP₃. Binding of IP₃ to its receptor on the endoplasmic reticulum (ER) results in a release of Ca²⁺ from this storage organelle to the cytosol. A depletion of Ca²⁺ in the ER triggers an opening of store-operated Ca²⁺ channels in the plasma membrane (to the far right and left), and a second wave of Ca²⁺ enters the cytoplasm. The Ca²⁺-dependent PKC may be activated by a combined effect of Ca²⁺ and DAG formed during hydrolysis of PIP₂.

Due to the large homologies at the amino acid level between the different G proteins, it has long been hard to determine with certainty the precise identity of the G protein involved in signaling, and bacterial toxins have basically been the only tools available. Bacterial toxins can be used to determine the involvement of Gαi and Gαs (inhibited by toxins from Bordetella pertussis and Vibrio cholera bacteria, respectively) whereas new types of inhibitors with specificity for the other G protein subunits have been eagerly awaited. Selective inhibitors of Gαq have recently been described,^16,17 and their use as inhibitors of neutrophil activation have revealed the following: (i) some neutrophil receptors use the Gαq-pathway to initiate a rise in [Ca²⁺]_i, whereas other (possibly Gαi-linked) use the Gβγ subunit (Figure 2), and (ii) the effects of pertussis toxin are not as selective as earlier anticipated (see section 2.1 and ref (12)).

1.4. Termination, β-Arrestin Binding, and Biased Signaling

Proteins belonging to the β-arrestin family of regulators that interact with GPCRs from the cytosolic side of the plasma membrane play important roles to dampen and terminate signaling induced through the G protein downstream of the agonist-activated GPCRs. Signaling of the agonist-occupied receptor is blocked for further signaling through the G protein, a receptor state termed homologous receptor desensitization. Yet, the recruited β-arrestins have modulating functions in signaling.¹⁸ The recruitment of β-arrestin from the cytosol and binding to the activated GPCR causes a physical separation of the actively signaling receptor/ligand complex from its G protein. The receptor bound β-arrestin can then directly bind to the adaptor protein 2 (AP2) complex and the heavy chain of clathrin¹⁹ which then initiates downstream signaling effects. In many cells, β-arrestins terminate parts of the signaling from the G protein and allow an activation of the extracellular-signal-regulated kinases or mitogen-activated protein kinases (ERK)-signaling cascade that secondary regulate other downstream functions.²⁰ Termination/desensitization of the G protein transduced signaling activities induced by GPCRs can thus be obtained through a direct physical separation of the G protein from the agonist-occupied receptor in the plane of the receptor expressing membrane. However, it is clear that other molecules can also bind to the receptor and be the basis for this type of physical separation of the G-protein from the receptor, and regarding neutrophils, actin in its polymerized form is one prominent example (Figure 3).²¹⁻²⁵

Figure 3.

Schematic drawing illustrating the transient rise in [Ca²⁺]_i and superoxide anion production by FPR1- and P2Y₂R-activated, -desensitized, and -reactivated neutrophils. (A) Activation of naïve neutrophils by FPR (left) and P2Y₂R (right) agonists, respectively, in the absence and presence of a cytoskeleton disrupting drug such as latrunculin A. In naïve neutrophils, FPR agonists induce a transient rise in [Ca²⁺]_i and an activation of the O₂^–-producing NADPH-oxidase (left), whereas a P2Y₂R agonist activates neutrophils without any generation of O₂^– (riht). Following a disruption of the cytoskeleton (Phase I), the level of O₂^– production induced by FPR agonists becomes increased (left), and a P2Y₂R agonist induces an activation of the NADPH-oxidase (right). (B) The response induced by FPR (left) and P2Y₂R (right) agonists, respectively, is rapidly terminated and the receptors are homologously desensitized; termination/desensitization of FPRs is achieved through a binding to the actin cytoskeleton of the receptor–ligand complex (left), whereas the mechanism responsible for the termination/desensitization P2Y₂R is unknown (right). (C) A disruption of the cytoskeleton in neutrophils with their FPRs desensitized by an FPR agonists (left) allows the receptors to be reactivated and procuce O₂^–, and this activation (Phase II) is achieved without any rise in [Ca²⁺]_i. Desensitized P2Y₂Rs (right) are not reactivated upon a disruption of the cytosleleton.

It should be noted that following the increased knowledge about the role of β-arrestins, the view of the downstream signaling cascade induced by GPCRs activated by an agonist has lately also been challenged; the precise conformational change in the receptor induced by the bound agonist determines the profile of the signal-pathways activated.²⁶ This type of unbalanced signaling has been termed “functional selective” or “biased signaling” and was initially described for two agonists that bind to the same receptor with the signaling routes initiated for example by 5-hydoxy-tryptamine (serotonin) receptor (5-HTR).²⁷ This concept is now known to be valid for many receptor/ligand pairs and thus implies that the classical linear two-state model, suggesting that receptors exist in an equilibrium between an inactive (off conformation) and an active (on conformation) state, and that this equilibrium is shifted toward the active state by agonist binding, is an oversimplification. Instead, GPCRs can adopt multiple conformations with different signaling properties distinctly induced by agonists with different receptor binding patterns. Since β-arrestins were recognized as proteins that are multifunctional, that is, mediating receptor desensitization and GPCR internalization and acting as scaffolds for adaptor and signal transduction proteins,²⁸ the identified number of ligands that trigger functional selective activities have steadily increased. Signaling that is functional selective/biased often discriminates between those that are β-arrestin-modulated and G-protein-dependent, respectively. This is, however, a notable simplification not only based on the fact that an observed difference does not necessarily have to be absolute as they might relate to efficacy differences for distinct signaling pathways.²⁹ Perhaps more important is that signaling activities regarded as downsteam of/modulated by β-arrestin may still rely on a functional G protein.³⁰⁻³² It should also be noted that the conclusion from an attempt to summarize and critically examine the different published data on the receptor downstream signaling pathways was that the roles in signaling of β-arrestin and G-proteins are distinct and non-overlapping and that β-arrestin binding to the activated receptor facilitates and redirects signaling to new downstream pathways.³³ It is thus clear that the assumption that receptors do not only exist in a dichotomy state of being either active or inactive but also can instead adopt various conformations is not only the basis for functional selectivity but also further extends this concept to a multiple-state model. In this model, any given ligand stabilizes a receptor conformation that is unique, and the number of possible conformations is only limited by the number of ligands which differ from each other in their efficacy to promote receptor interaction with available G proteins, G protein subtypes, and β-arrestin/actin.^26,34 Consequently, every agonist causes unique receptor conformations linked to specific signaling pathways and specific cellular responses.

1.5. Allosteric GPCR Modulation

By definition, allosteric GPCR-specific modulators transfer naïve nonsignaling receptors to a state that when activated gives rise to an increased (positively modulated) or decreased (negatively modulated) response upon ligand binding to the orthosteric site. It is thus clear that signaling by GPCRs is not determined by orthosteric agonists alone but also may be regulated/changed (modulated) by ligands having an alternative binding site. Such ligands or modulators are molecules that interact with a receptor at a location distinct from the orthosteric binding site. In general, this interaction occurs without inducing any activation, but the ligand induces an allosteric change of the targeted receptor and a sensitization or restraint response to conventional agonists. A low level of signaling by an orthosteric agonist can be sensitized by a nonactivating allosteric modulator that allows the orthosteric agonist to induce a stronger response (Figure 4, showing the basic model for an allosterically modulated response).³⁵ It is clear that the structural change of the receptor may alter the orthosteric binding site and/or the signaling capacity of the activated receptor. In principle, although there are exceptions, allosteric modulators are inert in the absence of an orthosteric agonist and do not fully compete in binding to the receptor of natural ligands or conventional antagonists. Synthetic receptor-binding compounds, that is, small molecules and endogenously produced molecules, may function as allosteric modulators. Some prominent GPCRs characterized for being influenced by endo- and/or exogenous allosteric modulators include 5-HT serotonin receptors, adenosine and adrenergic receptors, various chemokine receptors and opioid receptors, muscarinic acetylcholine receptors (mAChRs), dopamine receptors, and the GABA_B receptor.^36,37

Figure 4.

Activation and allosteric modulation of GPCRs. (A) An orthosteric GPCR-specific agonist transfers a naïve nonsignaling receptor to an active signaling state. For many GPCRs, a transient rise in [Ca²⁺]_i is an early receptor downstream signaling event. (B) A receptor-specific allosteric modulator binds to its receptor at a site distinct from the orthosteric binding site, and such ligands are inert in the absence of an orthosteric agonist. No active signaling such a transient rise in [Ca²⁺]_i is induced by an allosteric modulator, but the ligand induces an allosteric change of the targeted receptor. In principle, allosteric modulators are inert in the absence of an orthosteric agonist. (C) An allosteric modulator sensitizes its receptor, and the structural change of targeted receptor alters the response induced by a conventional agonist that binds to the orthosteric binding site. The transient rise in [Ca²⁺]_i induced by an orthosteric agonist is more pronounced when it binds to an allosterically modulated receptor, compared to when it binds to a naïve (nonmodulated) receptors (as compared to A).

1.6. Prototype GPCR and Neutrophil Pattern Recognition Receptors

The state-of-the-art model for how the regulatory machinery determines agonist mediated GPCR signaling and downstream functional output has for long been the human β₂-adrenergic receptor (β₂AR).^5,38,39 It is, however, clear that other members of the GPCR family may have different and sometimes more complex regulatory machineries as well as signaling behaviors. Thus, the cell type in which a particular receptor is expressed influences not only signaling but also the functional outcome of agonist/antagonist binding. Peptides with a formylated methionyl in their N-terminus, a hallmark of proteins/peptides synthesized by bacteria and mitochondria, were already in the mid 1970s shown to be danger molecules, chemotactic for human polymorphonuclear leukocytes (neutrophils), and they were at the time suggested to interact with a specific cellular binding site defined as a receptor.⁴⁰ This suggestion turned out to be correct, and the receptor is now known as formyl peptide receptor 1 (FPR1).⁴¹ Regarding the functions triggered, the chemotactic formyl peptides typically affect not only migration but also secretion of enzymes with proteolytic activities, activation of an NADPH-oxidase (a specialized electron transporting system that produces superoxide anions), and mobilization of receptors and adhesion molecules from intracellular storage granules,.^42,43 The N-formylmethionyl peptide-recognizing receptor (FPR1) was the first neutrophil receptor to be cloned^44,45 and shown to belong to the family of GPCRs. Even if two other closely related receptors (FPR2 and FPR3) as well as a large number of other neutrophil GPCRs have been identified and characterized since the cloning of FPR1, this receptor has become the prototype chemoattractant receptor in neutrophils. It is, however, clear that other neutrophil pattern recognition GPCRs such as the purinergic receptors (P2YRs) and FFARs have different and sometimes unique receptor downstream signaling machineries that affect and distinguish the associated functional repertoires of these receptors from those of both β₂AR and FPR1.^41,46

2. Signaling Downstream of GPCRs That Selectively Regulate Neutrophil Functions

2.1. Early Signaling Events Transduced in Neutrophils by Heterotrimeric G Proteins

When triggered by G proteins of different subtypes, the signaling pathways downstream of GPCRs can overlap, and the precise functional responses mediated by GPCRs depend highly on the individual receptor as well as on the ligand used to initiate the signaling process. The transient rise in [Ca²⁺]_i is one of the earliest events of receptor activation. The starting point for this signaling pathway is an activation of phospholipase C (PLCβ) downstream of the activated receptor/G-protein complex. Normally signaling by the G proteins containing a Gα_i subunit primarily affect the adenylyl cyclase activity and reduces the levels of cAMP in the cell through a response transduced by the Gαi subunit, but in neutrophils, the βγ subunit linked to Gαi is the part that activates PLC and thus regulates the release of Ca²⁺ from intracellular Ca²⁺ stores as well as the influx through SOCs in the plasma membrane.⁴⁷ The chemoattractant receptors FPR1 and FPR2 are prominent examples of Gα_i coupled neutrophil GPCRs, but the PLC–PIP₂–IP₃ signaling pathway is also initiated by the Gα subunit part in the class of G proteins containing a Gαq subunit. The purinergic receptor P2Y₂R that recognizes adenosine triphosphate (ATP) and the receptor for the lipid chemoattractant PAF (PAFR) are prominent examples of Gαq coupled neutrophil GPCRs.¹² The large sequence homologies between the different G proteins makes it hard to conclusively determine the identity of the precise G protein involved in signaling, and as mentioned, data obtained with bacterial toxins to determine the involvement of Gαiand Gαq are not always reliable. The reason for this statement is that despite the fact that the PAFR and the FPRs couple exclusively to Gαq and Gαi, respectively, pertussis toxin (that should affect only Gαi) inhibits signaling and cellular activation of both these receptors.¹² It is thus well-established that functions triggered by all neutrophil chemoattractant receptors, including the Gαq coupled PAFR, are inhibited by pertussis toxin,¹² and a proper dissection of the link between Gαi and different neutrophil GPCRs will be possible first when small inhibitors also selective for Gαi become available.

2.2. Lack of an Association between a Rise in [Ca²⁺]_i and Activation of the Neutrophil NADPH-oxidase

Neutrophils contain several types of intracellular membrane enclosed storage granules/vesicles that are formed during different developmental stages in the bone marrow.⁴⁸ These vesicle/granule membranes contain a large number of different adhesion molecules and chemoattractant receptors that are mobilized to the cell surface when naïve neutrophils enter a primed state and generally (but not always)⁴⁹ when the cells are recruited from the bloodstream to an inflamed tissue.⁵⁰⁻⁵² With respect to the role of Ca²⁺ in granule secretion, it has been shown that mobilization of the different granule/vesicle subsets is achieved in a strict rank order upon an elevation of [Ca²⁺]_i, with the easily mobilized secretory vesicles being exceptionally sensitive.⁴⁸ It is, however, clear that the granules are mobilized also in the absence of a transient rise in [Ca²⁺]_i as well as when receptor signaling is biased away from the PLC–PIP₂–IP₃ signaling route.^53,54 Taken together, current data suggest that an elevation of the [Ca²⁺]_i may induce or facilitate granule mobilization/secretion, but alternative pathways that trigger secretion independent of a change in [Ca²⁺]_i are also activated either alone or in parallel with the PLC–PIP₂–IP₃ signal transduction pathway.

Recruitment of neutrophils from blood to tissues by locally produced chemoattractants is a cardinal early event in inflammation, and the attractants typically stimulate the rate as well as direction of migration and switch neutrophils from a naïve or resting state, to a preactivated, primed phenotype; this process is in most (but not all)⁴⁹ cases associated with alterations in the composition of plasma membrane.^50,51 The ability of neutrophils to respond chemotactically has been shown to be intact not only in the presence of a Ca²⁺ chelator but also when the cytosolic concentration is substantially reduced/buffered,⁵⁵ suggesting that neutrophil chemotaxis is not dependent on transient changes of [Ca²⁺]_i. It should also be noted that in contrast to the residing view regarding tissue recruited neutrophils from skin models, neutrophils recruited in vivo as well as in vitro by synovial fluid derived chemoattractants, are basically resting, and with respects to the surface marker profile, they are in fact very similar to naïve cells isolated from peripheral blood,⁴⁹ showing that guiding of neutrophils by chemoattractant(s) may also operate without any granule mobilization.

The electron-transporting NADPH-oxidase present in neutrophil is an enzyme system that comprises five protein subunits: two that are membrane-bound and form a b-type cytochrome heterodimer and three that are present as a complex in the cytosol of nonactivated cells. Upon activation, the cytosolic factors are translocated to the b-cytochrome containing membranes (i.e., the plasma membrane and granule membrane) where they assemble and form a functional enzyme system.^56,57 The assembled and activated NADPH-oxidase facilitates the generation of superoxide anions (O₂^–) generated through transport of electrons from NADPH in the cytosolic across the membrane to molecular O₂ that is reduced on the opposite side of the membrane; the O₂^– produced may dismutate to hydrogen peroxide (H₂O₂), and several other reduced oxygen species. A large number of stimuli, including many GPCR agonists, activate neutrophils to produce O₂^–, but as yet, not all the signaling pathways with regulatory roles in the activation processes have been identified. It is, however, well-established that there is no direct link between an activation of the superoxide generating NADPH-oxidase and a transient rise in [Ca²⁺]_i, illustrated by the following facts: (i) the ATP/P2Y₂R agonist/receptor complex induces a transient rise in [Ca²⁺]_i but no activation of the oxidase,²² (ii) the NADPH-oxidase activation, triggered by desensitized FPRs when reactivated through a disruption of the cytoskeleton, occurs without any parallel rise in [Ca²⁺]_i,²¹ (iii) the receptor cross-talk signals generated by PAFRs that reactivate desensitized FPRs also occurs without any parallel rise in [Ca²⁺]_i,⁵⁸ and (iv) Activation of the NADPH-oxidase by interdependent allosteric free fatty acid 2 (FFAR2) modulators occurs without any transient rise in [Ca²⁺]_i.⁵⁹

2.3. β-Arrestin Binding and Modulation of Signaling

The β-arrestin proteins play important roles in GPCR signaling, and two of these (arrestins 2 and 3) that are expressed basically in all cells have functionality for many receptors to stop signaling that is dependent on an activated G protein downstream of receptors with a bound agonist. The nonsignaling β-arrestin occupied receptor is regarded as homologously desensitized.¹⁸ Such desensitized receptors may, however, transduce the signals leading to an uptake of agonist-occupied receptors through and endocytic process, and the bound β-arrestin proteins are also able to modulate the further downstream signaling by activated receptors. A changed level of receptor phosphorylation induced by activated G protein coupled receptor kinases (GRK), facilitates β-arrestin binding to receptor parts facing the cytosol, and this binding sterically/physically blocks binding of the cognate G protein to the receptor. The endocytic process is initiated when β-arrestins bound to a receptor recruit the AP2-complex together with the heavy chain of clathrin, a recruitment that leads to a clustering of receptors and formation of clathrin-coated pits.⁶⁰ To complete the endocytic process the coated-pits have to be cleaved off from the plasma membrane, and this fission process requires a membrane remodelling enzyme (a GTPase dynamin). The receptors present in endocytic vesicles are either degraded, or if the binding of β-arrestin is weak and transient, the receptors may be recycled back to the plasma membrane.

It has also been suggested that when β-arrestins are bound to an agonist activated receptor they can not only recruit but also actually directly interact with (and possibly activate) other proteins with signaling capacities such as Src kinases and ERK1/2; the signals generated by these are important in both receptor desensitization and endocytosis. It has also been shown that receptor bound β-arrestins affect the activity of regulatory ubiquitin E3 ligase and a stress-related JNK3 kinase.⁶¹⁻⁶³ Much is still unknown about the role of β-arrestin in signaling, and it might be that the modulatory fine-tuning determining the functional outcome of GPCR responses may require the involvement of an initiating G-protein.⁶⁴ This would then be the starting point for interactions with other proteins and distinct conformational changes in the receptor and cytosolic proteins (including β-arrestin) recruited to the plasma membrane.

2.4. β-Arrestin and Lessons from Studies of Neutrophil Chemotaxis

As mentioned, β-arrestins have in many cells a role as adaptor molecules that when bound to an agonist-occupied receptor both inhibit the contact physically, between the receptor and the signaling G proteins, and also shift the signaling cascade to ERK-regulated pathways/functions, including the removal from the cell surface of the agonist-occupied receptors.⁶⁵ The role of β-arrestins in modulating neutrophil GPCR signaling is, however, more complex, as illustrated by the fact that FPR2 agonists unable to recruit β-arrestin activated ERK as efficient as those that recruit β-arrestin⁶⁶ and that when the protein is overexpressed in different cell lines ERK is rapidly and transiently activated by the signals generated downstream of activated FPRs. This signaling is most probably mediated through an activated G protein.⁶⁷ Moreover, it has been shown that internalization of FPR2 but not of FPR1 also requires an expression of β-arrestin, but there is no such requirement for endocytic uptake of occupied receptors.⁶⁷ It should also be noted that it has been suggested that lipid FPR2 agonists have the capacity to recruit β-arrestin, however without any direct functional outcome in neutrophils, but these data have been questioned.^46,68−70 Furthermore, recent data from the researchers that suggested that lipid FPR2 agonist recruit β-arrestin do not confirm their earlier published results,⁷¹ suggesting that FPR2 is not the receptor for these lipid mediators (see also ref (46)). With respect to the role of β-arrestin-modulation of signaling important for granule secretion/mobilization and subsequent upregulation of adhesion molecules to the plasma membrane, data have demonstrated this to be achieved equally well by FPR-specific agonists that recruit β-arrestin, and those that do not.⁷²

Chemotactic recruitment of neutrophils from blood to tissues is a cardinal event in inflammation, and in in vitro settings, FPR2 agonists unable to recruit β-arrestin are also unable to initiate directional migration.^66,72,73 It remains, however, to determine if there is a direct link between β-arrestin recruitment and chemotaxis, and future identification of tool compounds that selectively trigger β-arrestin and engineering of motile cells that are deficient in β-arrestin should help gain more insights into this aspect.

In neutrophils, the NADPH-oxidase is assembled and activated to produce O₂^– as well as terminated both by FPR2 agonist that recruit and are unable to recruit β-arrestin.^66,72,73 This imply that activation and termination the signaling pathways that regulate O₂^– production are independent of β-arrestin recruitment

3. Neutrophil GPCRs and Regulatory Functions of the Actin Cytoskeleton

3.1. Actin Cytoskeleton Plays a Role in Termination of Signaling and in Homologous Receptor Desensitization

Signaling activities downstream of activated GPCRs can be terminated through a physical separation in the plasma membrane of the signaling G protein and the agonist-occupied receptor. It is clear that for many receptors β-arrestin is the adaptor molecule that is recruited to the cytosolic domains of a signaling GPCRs, and its binding then physically blocks the contact between the G protein and the receptor.⁷⁴⁻⁷⁶ The role of β-arrestins in signaling by the neutrophil pattern FPRs is, however, more complex. The agonist-occupied FPRs rapidly and transiently activate the ERK-signaling pathway, and this activation pathway is induced also by receptor-specific agonists that lack the ability to recruit β-arrestin.^66,72 These results are in agreement with fact that ERK has been shown to be a downstream target of the activated G protein and not downstream of β-arrestin,³⁰⁻³² but it can possibly also be independent of both β-arrestin recruitment and G protein activation.⁶⁶ Signaling by activated FPRs is also rapidly terminated when the activating agonists are unable to recruit β-arrestin, and these receptors become nonresponsive (desensitized) to other agonists recognized by the same receptor.^66,72 There are thus no direct links between β-arrestin and homologous receptor desensitization or β-arrestin and ERK activation in neutrophil GPCR signaling, and this is in line with the notion mentioned above, that that ERK activation depends on G proteins rather than on β-arrestin recruitment.³⁰⁻³²

In naïve neutrophils, the NADPH-oxidase derived superoxide anions generated by downstream signals from agonist-occupied FPRs, are produced for a fairly short time period, and following the rapid termination of this signaling, the receptors become homologously desensitized.⁷⁷ Cytoskeletal proteins (polymerized actin and/or different actin binding proteins linked to the cytoskeleton) in the actin-rich cortex-lining on the cytosolic side of the plasma membrane, rather than β-arrestin, have in neutrophils been shown to constitute the basis for receptor interactions leading to termination of this signaling (Figure 3A, left).²⁴ The receptor-bound polymerized actin thus constitutes the basis for the termination/desensitization, a state possibly achieved through a physical separation of the G protein from the activated receptor (Figure 3B, left).²⁴ It is clear that these interactions are reciprocal, meaning that signaling membrane receptors affect cytoskeletal functions, but at the same time, the cytoskeleton regulates receptor functions through interactions with lipid components and/or membrane proteins. The involvement of the actin cytoskeleton in the termination/desensitization process in neutrophils is evident from experiments in which the integrity of the cytoskeleton is disrupted by drugs that inhibit actin polymerization. Such a disruption of the cytoskeleton affects desensitization of the agonist-occupied FPRs and modulate the receptor-mediated neutrophil responses.^78,79 In accordance with the role of the cytoskeleton in the termination/desensitization, the response triggered by FPRs and many other neutrophil GPCRs in neutrophils lacking an intact cytoskeleton, is both increased in magnitude and largely prolonged.^23,25,52 This signaling pattern is valid not only for receptor agonists that trigger β-arrestin recruitment but also for those that activate the receptor without such a recruitment.^66,72 Despite the fact that FPRs are desensitized by mechanisms not involving β-arrestin, it has been shown that neutrophil-like HL-60 cells lacking β-arrestin have reduced capacity to migrate chemotactically in response to FPR agonists.⁸⁰ This indicates that chemotaxis relies on signals generated downstream receptor-recruited β-arrestin, and this suggestion gains support also from the fact that non-β-arrestin-recruiting FPR2 agonists are functionally biased away from chemotaxis.^66,72 It is clear from earlier studies that β-arrestin-modulating activities and G protein-dependent signaling (and their physiological consequences) are not linked events,³⁰⁻³² and this is obviously true also in neutrophil GPCR-signaling. There seems, however, to be a link between β-arrestin recruitment and chemotactic migration.

As mentioned, the presence of drugs that inhibit actin-polymerization, increases the magnitude of the response induced by activated FPRs in neutrophils isolated from peripheral blood (Figure 3A, left). This effect is less pronounced in tissue neutrophils recruited from the bloodstream to an aseptic skin inflammation.⁵² The difference between these cells is that naïve blood neutrophils have the majority of their FPRs stored in intracellular granules, whereas these pools are largely lacking in tissue-recruited neutrophils, having mobilized their granules to the plasma membrane. The responses are largely prolonged by cytoskeletal-disrupting drugs in both cell populations (unpublished data), suggesting that the actin cytoskeleton plays a primary role in the termination process.

3.2. P2Y₂R-Mediated NADPH-oxidase Activation Is Blocked but Not Terminated by the Actin Cytoskeleton

Functional selectivity/biased signaling was initially thought to require different agonists binding to the extracellular part of a receptor. Yet, it is obvious that this type of selectivity can also be achieved by regulatory components that interact specifically with cytosolic receptor domains from inside the plasma membrane.⁸¹ It has for example been shown that cytosolic antibodies directed toward cytoplasmic domains of the β₂AR, selectively and from inside, block one signaling pathway downstream of this receptor,⁸² and it is clear that in neutrophils the cytoskeleton may also convey such a functional selectivity. The nucleotide ATP is a danger-associated molecular pattern (DAMP) molecule released from injured tissues/cells, and this signaling molecule is recognized by purinergic GPCRs.^83,84 Activation of the neutrophil ATP receptor P2Y₂R induces a transient rise in [Ca²⁺]_i. However, contrary to the response triggered by FPR agonists, the signals generated by P2Y₂R are functionally selective; no activation of the superoxide producing NADPH-oxidase occurs in parallel to the rise in [Ca²⁺]_i.²² The ATP-occupied P2Y₂Rs mediate, however, oxygen radical production when the cytoskeleton has been disrupted (Figure 3A, right). That is, in neutrophils with a dynamically intact cytoskeleton, the P2Y₂R signals supposed to activate the oxidase are blocked, and this blocking is mediated by an actin-cytoskeleton-dependent mechanism.²² The cytoskeleton thus conveys functional selectivity of P2Y₂R, a function regulated from the cytosplasmic side of the neutrophil plasma membrane. Hence, the state of the cytoskeleton thus determines the signaling route downstream of the ATP/P2Y₂R agonist-receptor complex, and a new signaling route becomes available for the agonist (i.e., ATP)-occupied P2Y₂Rs. The quality of the downstream signals is regulated from inside the plasma membrane.

It is thus clear that agonist binding to P2Y₂R activates downstream signals/effector functions that differ depending on whether the cytoskeleton is intact or disrupted. Irrespective of the status of the cytoskeleton, the signals downstream the ATP-occupied P2Y₂Rs rely on a functional Gα_q-protein.⁸⁵ In naïve neutrophils, the ATP/P2Y₂R/Gα_q complex thus activates the signaling route that leads to an increase in the cytosolic concentration of free Ca²⁺, whereas the signaling route leading to an activation of the NADPH-oxidase is blocked (Figure 3A, right). That is, different receptor signals triggered by one single receptor agonist when bound to its receptor can be changed by the cytoskeleton, and this regulatory machinery is present inside the plasma membrane of the receptor-expressing cell.²² The molecular mechanism for this modulation, by which the signaling properties of an activated GPCR can be blocked or switched to a new type of functional selective response remains to be elucidated. In-depth knowledge about this regulatory mechanism will possibly facilitate the design of novel drugs affecting innate immune reactivity and/or inflammatory processes and increase the understanding of GPCR signaling in general.

The data on P2Y₂R-triggered signaling also implies that this receptor uses a cytoskeleton-independent mechanism to terminate the neutrophil response,²² a mechanism that remains to be defined and further characterized (Figure 3B, right). The suggestion of a cytoskeleton-independent termination mechanism, is based on the assumption that neutrophil activation by FPR agonists and ATP, of cells in which the cytoskeleton is disrupted prior to agonist binding (allowing for the activated P2Y₂R to access the signaling route leading to an activation of the NADPH-oxidase), should be very similar, but this is not really the case. Superoxide production is turned off much more rapidly in ATP-activated neutrophils than when an FPR agonists is used as the activating ligand. In fact, the P2Y₂R-induced superoxide response in neutrophils with a disrupted cytoskeleton, resembles that induced by agonist-occupied FPRs in neutrophils in which the cytoskeleton is dynamically and functionally intact. That is, also with a disrupted cytoskeleton, neutrophils are equipped with a system that rapidly terminates the P2Y₂R triggered response.²² It is obvious that although the cytoskeleton actively blocks signaling by P2Y₂R the FPR model for termination of signaling is not applicable to the P2Y₂R response, and an alternative, non-cytoskeleton-dependent alternative mode, must be put in action.

3.3. Actin Cytoskeleton Regulates Receptor Reactivation of Some but Not All Desensitized Neutrophil GPCRs

The desensitized receptor state has long been regarded as irreversible, meaning that such receptors should not be available for signaling until the internalization/recycling process has been accomplished and the receptor is circulated back to the cell surface. This unavailablity for signaling is, however, not valid for desensitized FPRs as a disruption of the cytoskeleton reactivates these desensitized receptors, and moreover, the signaling downstream of the reactivated desensitized receptors is biased. As opposed to direct FPR activation, the reactivation signals generated upon disruption of the cytoskeleton, activate the radical producing NADPH-oxidase but yet this is achieved without any activation of the PLC–IP₃–Ca²⁺ signaling route (Figure 3C, left;²¹). In contrast, a disruption of the integrity of the actin cytoskeleton does not generate any activation signals by the agonist-occupied and desensitized PAFRs, P2Y₂Rs and FFARs (Figure 3C, right),^22,52,53 despite the large increase in respiratory burst activity induced by agonist specific for these receptors, in neutrophils in which the cytoskeleton is disrupted prior to receptor activation through addition of the agonist. The effect of cytoskeleton disruption is lost very rapidly after receptor activation for example with ATP, suggesting that the non-cytoskeleton-dependent desensitization of P2Y₂R is put into action promptly after agonist binding. The cytoskeleton might serve as a fail-safe mechanism for neutrophils to limit P2Y₂R-induced oxygen radical production and prevent unnecessary damage to injured tissues. Hypothetically, there might be endogenous, not yet identified P2Y₂R ligands that have the capacity to also induce an activation of the NADPH-oxidase in naïve neutrophils having an intact cytoskeleton that blocks the activation signal when ATP is the activating agonist (see also the receptor cross-talk activation mechanism described below).

4. Pepducins as Regulators of GPCRs in Neutrophils Receptor Hijacking and Activation/Inhibition Similarities to Orthosteric Ligands

4.1. Activation of Human Neutrophils by a Pepducin with an Amino Acid Sequence Identical to the Third Intracellular Loop of FPR2: The Specificity Part of the Pepducin Concept Is Fulfilled

A unique class of ligands for GPCRs, specifically designed lipopeptides (pepducins), was described some 15 years ago. It was shown that these GPCR ligands regulate receptor activity through a unique mechanism and a novel concept for ligand–receptor–interaction was introduced (Figure 5). Pepducins are synthetic lipopeptides containing a fatty acid (usually palmitate) coupled to a short peptide chain (usually 10–20 amino acids). The amino acid sequence in the pepducin is identical to one that is present in one of the intracellular domains (the cytoplasmic tail or one of the three intracellular loops) of the GPCR that should be targeted. To be able to modulate receptor function, pepducins are supposed to use the fatty acid part to transverse the plasma membrane, and once inside, the peptide part is expected to interact with the signaling parts of the receptor and possibly also the signaling G protein. The interaction is supposed to take place on the cytoplasmic side of the membrane. That is, the GPCR/pepducin binding is presumed to take place on the on the inside of the plasma membrane and involve receptor parts facing the cytosol.⁸⁶⁻⁸⁸ This type of interaction is in obvious contrast to the binding/signaling characteristic of conventional GPCR ligands that mediate their function through a direct interaction with orthosteric binding sites present on receptor parts exposed on or close to the cell surface. The suggested mode of action for pepducins is unique: Once the peptide part of the pepducin has transversed the plasma membrane to the cytosolic side of the membrane the pepducin interacts with parts of the receptor and/or the signaling G protein, and the receptor specificity should be discerned by the sequence identity/similarity between the pepducin and the GPCR that should be selectively targeted.^89,90 The result of such an interaction between two identical peptide sequences present in the lipopeptide and the targeted receptor, respectively, will be either an inhibition or an activation of receptor functions.⁸⁶ The basic specificity concept has been shown to be valid for one of the neutrophil FPRs; FPR2 is activated by a pepducin containing a peptide chain having the same amino acid sequence as the third intracellular loop of FPR2.⁹¹ It is thus assumed that pepducins act through an interaction with corresponding intracellular region of the specifically targeted GPCR, and the effects are achieved through an allosteric modulation of intracellular receptor domains that couple to G proteins and other signaling/regulating proteins. However, data obtained with pepducins designed to interact with different neutrophil GPCRs suggest that there might be different, and sometimes possibly unique modes of action related to each individual receptor/pepducin pair (Table 1). This is clearly illustrated by the results obtained with F2Pal₁₀, a palmitoylated peptide with an amino acid sequence identical to sequence present in the third intracellular loop of FPR2. The F2Pal₁₀ pepducin selectively activates FPR2 over the closely related FPR1, but in disagreement with the concept, F2Pal₁₀ also activates a mutated FPR2 in which two amino acids have been replaced giving rise to a chimeric receptor with an amino acid sequence in the third intracellular loop identical to that in FPR1.⁹¹ Moreover, F2Pal₁₀ also activates a chimeric receptor in which the cytoplasmic tail of FPR1 has been replaced by that of FPR2, leaving the third intracellular loop of FPR1 intact.⁹² These data, showing that there is no direct link between the amino acid sequence in the activating pepducin and that of the receptor mediating the response, suggest that at the receptor level sequences/regions apart from those in the third intracellular loop is involved in pepducin sensitivity.

Figure 5.

Mode of action for GPCR specific pepducins. The established concept for interaction of pepducins (lipopeptides with a peptide containing a sequence of amino acids identical to that in the receptor to be targeted; left) is that the ligand anchors to the cell membrane through its fatty acid moiety (usually palmitate), and when the peptide part of the molecule has been transferred (flipped) to the cytosolic side of the membrane, it presumably interacts with receptor domains exposed to the cytosol. This interaction initiates an inside-in signaling. In principal, conventional agonists and activating pepducins should induce very similar receptor-dependent cellular responses. Whereas a response induced by an orthosteric agonist is inhibited by a receptor-specific antagonist, the response induced by the pepducin is insensitive. The basic criteria for the pepducin concept are not met in neutrophils as different pepducins target FPR2 despite the fact that they are designed to target other receptors. FPR2 activating/inhibiting pepducins (i) affect receptor functions also without any similarities between the amino acid sequence of the pepducin and the cytosolic receptor parts, and (ii) pepducin-induced FPR2 activities are inhibited by conventional receptor specific antagonists. The data available suggest that FPR2-selective pepducins regulate receptor function through an interaction with the orthosteric binding site rather with intracellular parts of the GPCR.

Table 1. Characteristics of Neutrophil-Modulating Palmitate-Containing Pepducins.

pepducin name	receptor origin	peptide sequence and length	expected target/effect	target and effect in neutrophils	ref
F1Pal16	FPR1	third ICL K227–V242 (16 aa)	FPR1 (no effect)	FPR2 antagonist/Fpr2 agonist	(95,98)
F2pal16K5→Q	FPR1/2	third ICL K227,K231→Q–V242 (16 aa)	FPR1/FPR2 (active)	FPR2 antagonist	(91)
F2pal16M7→L	FPR1/2	third ICL K227,M233→L–V242 (16 aa)	FPR1/FPR2 (active)	FPR2 agonist	(91)
F2Pal16	FPR2	third ICL K227–V242 (16 aa)	FPR2 (active)	FPR2/Fpr2 agonist	(91,98,117)
F2Pal10	FPR2	thirdICL K227–S236 (10 aa)	FPR2 (active)	FPR2/chimeric receptor agonista	(92,113)
ATI-2341	CXCR4	first ICL M63–L78 (16 aa)	CXCR4 (no effect)	FPR2 agonist	(97,118)
P2Y2 PalIC2	P2Y2R	second ICL H130–R152 (23 aa)	P2Y2R (no effect)	FPR2 agonist	(93)
P2Y2 PalIC3	P2Y2R	third ICL M221–T246 (26 aa)	P2Y2R (no effect)	FPR2 agonist	(93)
P4 Pal10	PAR4	third ICL S274–R283 (10 aa)	Gαq proteins (no effect)	FPR2 antagonist/FFA2R agonistb	(86,99)
mF1Pal16	Fpr1	third ICL K238–V253 (16 aa)	Fpr1(no effect)	Fpr2 agonist/FPR2 antagonist	(98)
mF2Pal16	Fpr2	third ICL K227–V242 (16 aa)	Fpr2 (active)	Fpr2/FPR2 antagonist	(98)

^aF2Pal₁₀ activates a mutated FPR2 in which two amino acids have been replaced in the third intracellular loop, making it identical to that in FPR1. F2Pal₁₀ also activates a chimeric FPR1/FPR2 in which the cytoplasmic tail of FPR1 has been replaced by that of FPR2.

^bRequires allosterically modulated FFAR2.

4.2. FPR2 Recognizes Many Pepducins Unrelated to This Receptor

The basic interaction characteristics in neutrophils are not always in agreement with the expected interaction/specificity profile, and this is valid for several different pepducins. It has also been shown that pepducins that lack effects on the expected receptor, based on amino acid sequence identity, have off-target effects through hijacking of FPR2.^91,93−95 The fact that FPR1 and FPR2 are very similar and differ in only two amino acids in the third intracellular loop⁹⁶ could possibly explain the hijacking of pepducins with an amino acid sequence identical to that in the third intracellular loop in FPR1. It is, however, clear that FPR2 is also hijacked by pepducins with peptide sequences designed for CXCR4 and P2Y₂R, receptors with no direct similarity to FPR2 at the amino acid sequence level.^93,97 In addition, and in disagreement with the pepducin concept, neutrophil activation induced by FPR2 pepducins is inhibited by conventional FPR2 antagonists that act at the extracellular surface of the receptor. The precise mechanism(s) for FPR2 hijacking by pepducins has not yet been elucidated but raises the question of whether FPR2 is a pattern-recognition receptor for some types of fatty acid coupled peptides.

4.3. Pepducins Recognized by Human FPR2 Are Recognized Also by Its Mouse Orthologue but Not Necessarily with the Same Modulatory Outcome

The pepducin technology/concept has also been used to determine ligand recognition profile similarities/differences between the FPRs in human neutrophils and their mouse orthologues.⁹⁸ The effects of two pepducins with amino acid sequences identical to that in the third intracellular loops of FPR1 (F1Pal₁₆) and FPR2 (F2Pal₁₆), respectively, have been determined and compared to the activities mediated by the corresponding pepducins from the mouse orthologues Fpr1 (mF1Pal₁₆) and Fpr2 (mF2Pal₁₆). In human neutrophils, the FPR pepducins selectively target FPR2, but with a different outcome, either inhibiting (F1Pal₁₆) or activating (F2Pal₁₆). It should be noted that F2Pal₁₆ activates not only this receptor but also Fpr2, a receptor crossover achieved despite the fact that 7 out of the 16 amino acids differ in the third intracellular loops (supposed to be the basis for selectivity) of FPR2 and Fpr2. The FPR1/Fpr1-derived pepducins (F1Pal₁₆ and mF1Pal₁₆) are very similar (differing in only one amino acid), and they also modulate neutrophil function similarly in the two species, not however through FPR1/Fpr1 but through FPR2/Fpr2. That is, both these pepducins are recognized by FPR2 and Fpr2, whereas they antagonize the activity triggered by orthosteric FPR2 agonists, they both directly activate the mouse variant Fpr2 (Table 1).⁹⁸

4.4. FPR Activation Characteristics Call the Pepducin Concept in Question

Pepducin are expected to selectively affect (positive or negative) functions only of the receptor that contains a peptide sequence exposed to the cytosol that is identical to that in the pepducin or as shown for P4 Pal₁₀ (see below) with several receptors having in common that signaling is transduced though a Gαq-containing heterotrimeric G protein.⁹⁹ For proper function, a pepducin has to be translocated to the inner leaflet of the membrane containing the receptor and then in a direct or indirect manner interact with an identical peptide sequence present in the cytosolic parts of the targeted receptor. This fact makes it hard to understand the basic molecular mechanism for how the peptide part of a pepducin can regulate receptor function specifically. Irrespective of the precise mechanism of action, the original model for how pepducins mediate their functions⁸⁹ contains some important inherent restrictions: (i) A pepducin should only affect signaling by receptors that contains an identical amino acid sequence in one of the intracellular signaling domains, with the protease-activated receptor 4 derived pepducin (P4 Pal₁₀) being an exception since this pepducin affect receptors that signal though a Gαq-containing heterotrimeric G protein.⁹⁹ (ii) Blocking of the surface-exposed binding site for orthosteric agonists with a conventional antagonist that simply blocks agonist binding should not have any impact on signaling induced by a receptor-activating pepducin. It is clear that the basic validation criteria for the pepducin concept are not met for pepducins designed to target several different neutrophil GPCRs. Several different pepducins designed for different receptors target the neutrophil FPR2, and the effects of these pepducins are inhibited by conventional FPR2 antagonists. In addition, it has been shown that pepducins derived from the second and third intracellular loops of P2Y₂R, the receptor for ATP, as well as one from the first intracellular loop of CXCR4 (AT-2341), the receptor for CXCL12 (stromal derived factor 1), hijack FPR2 and activate human neutrophil functions, data that add more complexity to how pepducins regulate GPCR signaling.⁹³ It is also clear that P4 Pal₁₀, the protease-activated receptor 4 derived pepducin, inhibits signaling downstream of several unrelated GPCRs having in common that they are Gαq-coupled,⁹⁹ but it lacks inhibitory effects on signaling downstream of the Gαq-coupled receptors for ATP (P2Y₂R) and PAF (PAFR) expressed in human neutrophils. However, P4 Pal₁₀ affects the function of neutrophil receptors that are Gαi-coupled; the pepducin inhibits neutrophils activated by orthosteric FPR2 agonists and activates the neutrophil FFAR2 when this is sensitized by an allosteric modulator.¹⁰⁰

Taken together, these results raise some questions of general character dealing not only with the molecular mechanism for specific GPCR activation by pepducins but also with recognition characteristics of FPR2 and FFAR2. It is, however, clear that not only the sequence of amino acids but also the position of charged amino acids (rather than the net charge) in a pepducin are characters of importance for function,^46,93,101 and this suggests that there are some strict structural requirements that determines if a pepducin is recognized by FPR2 or not.

5. Neutrophil FFAR2/GPR43 as Model System to Study Allosteric Receptor Modulation

5.1. Allosteric Receptor Modulation

With the identification of ligands lacking a direct effect on receptor signaling but found to potentiate (positive modulation) or inhibit (negative modulation) receptor functions induced by conventional agonists that trigger receptor activity through a direct interaction with the orthosteric binding site, an alternative concept for how GPCR signaling can be regulated was introduced (Figure 4).^29,39,102 According to the allosteric receptor modulation concept, an allosteric GPCR modulator binds to a distinct receptor site, separated from the part of the receptor that recognizes a true agonist or antagonist, and this binding occurs without any direct functional or signaling effects on the receptor. The interaction/binding of an allosteric modulator to a specific allosteric binding site instead transfers the receptor to (i) a state that has a lower energy barrier needed to induce the conformational change required for a transfer of a nonsignaling receptor to an activated signaling one, secondarily leading to an increased signaling activity downstream of the agonist-occupied receptor; (ii) a state in which the binding affinity for conventional (orthosteric) agonists is changed; or (iii) a state in which the two mechanisms are combined. By definition, the allosteric effect is reciprocal, meaning that a positive allosteric effect is achieved both when an orthosteric agonist activates an allosterically modulated receptor and when an allosteric modulator binds it is specific binding site in a receptor with it is orthosteric binding site occupied by an conventional agonist. Although, allosteric GPCR ligands have been described that have unprecedented modes of action and widen the allostery concept,¹⁰³ the basic selectivity restriction, stating that an allosteric receptor modulator affects the response triggered solely by orthosteric ligands that bind to the same receptor as the modulator (irrespective of whether this is a receptor monomer or homodimer), is not yet directly challenged by these data.^104,105 It should, in addition, be noted that there are ligands that display both agonistic and allosteric properties, and such ago-allosteric agonists may have an allosteric binding site that involves (overlaps with) the orthosteric binding pocket.¹⁰⁶

5.2. Allosteric FFAR2 Modulation Turns Orthosteric Agonists into Potent Neutrophil-Activating Ligands

The allosteric FFAR2 modulators, Cmp58, and AZ1729^107,108 prime neutrophils in their response to FFAR2 agonists (Figure 6A).^52,53,85 The mechanisms at a molecular level that underlie neutrophil priming has been studied extensively during the years, and several different mechanisms have been suggested to be of importance,⁴⁶ and now allosteric receptor modulation can be added to these mechanisms. It is clear that the allosteric modulators transfer natural FFAR2 agonists (i.e., acetate, propionate, and butyrate) that normally are non-/low-activating to potent activating agonists that trigger an assembly of the neutrophil superoxide generating NADPH oxidase, and at the same time, the threshold for the FFAR2 induced intracellular Ca²⁺ release is lowered. These data together constitute the basis for suggesting allosteric receptor modulation as a novel priming mechanism in neutrophils. The allosteric energy (the interaction between the two ligands through the receptor) is reciprocal, suggesting a consistent allosteric effect between the modulator and acetate/propionate. Irrespective of the precise modulation mechanism, the outcome of the modulating effect in neutrophils is that acetate/propionate is turned into agonists that activate neutrophils to generate and release superoxide anions. Short-chain fatty acids (SCFAs) are produced by gut microbes during anaerobic metabolism, and this suggests that receptors for such molecules expressed on innate immune cells are of prime importance for regulating inflammatory reactions in the gut.^109,110 However, the observation that tissue-recruited neutrophils to aseptic inflammatory sites are nonresponsive (possibly desensitized) to FFAR2/GPR84 agonists^52,53 also suggests a role for the receptor(s) in other organs and implies a functional complexity that needs further investigation.

Figure 6.

New models of GPCR activation through a receptor cross-talk mechanism and interdependent allosteric modulators, respectively. (A) According to the dogma, (i) no neutrophil activation (measured as a transient rise in [Ca²⁺]_i and O₂^– production) is induced by the allosteric FFAR2 modulators Cmp58 and AZ1729, (ii) the orthosteric agonist propionate partly activate neutrophils (an increase in [Ca²⁺]_i but no O₂^– production), and (iii) an increased response is induced by proipionate when combined with one of the allosteric FFAR2 modulators. (B) An allosteric modulator should affect receptor-mediated activities solely when triggered by an orthosteric agonists, but allosteric FFAR2 modulation by either Cmp58 or AZ1729 also positively modulates neutrophils in their response to the P2Y₂R specific agonist ATP. In accordance with a receptor cross-talk mechanism, the FFAR2 antagonist CATPB abolishes the ATP-induced activity, which verifies that FFAR2 is part of the Gαq-transduced response induced by ATP. The signaling by the allosterically modulated FFAR2 is not Gαq-transduced when propionate is the activating agonist, and ATP/P2Y₂R-mediated activation is obtained only when the FFAR2s are allosterically modulated. (C) No neutrophil activation is induced by either of the allosteric FFAR2 modulatiors Cmp58 and AZ1729 alone, but the two modulators together activate the cells. The activation by the two interdependent allosteric modultors is functionally selective; the activation of the neutrophil O₂^– generating NADPH-oxidase is not accompanied by a concomitant transient rise in [Ca²⁺]_i.

5.3. Novel Receptor Cross-Talk between GPCRs Highlighted by Allosteric FFAR2 Modulation

The generally accepted model for how an allosteric modulator affects receptor function is that the modulator binds to a specific allosteric binding site. Occupation of the allosteric binding site by a positive modulator transfers the receptor to a state that binds orthosteric agonist with higher affinity and/or have a reduced energy barrier for the conformational shift that transfers the receptor from a nonsignaling to an actively signaling state (Figure 6A).^29,111 According to this dogma, an allosteric GPCR modulator should then affect receptor-mediated activities solely when triggered by orthosteric agonists that are specific for this particular receptor. Allosteric modulation of FFAR2 positively modulates (primes) neutrophils in their response not only to FFAs that bind the orthosteric site of FFAR2 but also to ATP, a P2Y₂R specific agonist (Figure 6B).⁸⁵ This suggests that the allosteric receptor modulation–restriction concept is not by necessity a distinctive functional outcome that follows an allosteric receptor modulation. The allosteric modulatory effect is obviously not restricted to the orthosteric binding site of the modulated receptor. The mechanism suggested for how the signals generated by the agonist-occupied P2Y₂Rs can activate FFAR2, is that activation is achieved through an intracellular receptor cross-talk between the two receptors P2Y₂R and FFAR2: The activation signaling starts with a Gαq-containing G protein downstream of the activated P2Y₂R, and the signal(s) generated then activates the allosterically modulated FFAR2s from the cytosolic side of the membrane. In accordance with the suggested mechanism for how one GPCR can activate another receptor, the FFAR2 antagonist CATPB abolishes the response induced when ATP binds its receptor. This clearly shows that a participation of FFAR2 is required for the Gαq-transduced response induced by ATP in cells with their FFAR2s modulated. It is also important to note the following: (i) the ATP/P2Y₂R mediated activation is obtained only when the FFAR2s are allosterically modulated, (ii) the orthosteric agonist propionate is not able to replace the allosteric modulator, and (iii) signaling by the allosterically modulated FFAR2 is not Gαq-transduced when propionate is the activating agonist.⁸⁵ In addition, this receptor cross-talk modulation/activation mechanism by which the P2Y₂R agonist activates the allosterically modulated FFAR2 is not reciprocal, meaning that no activation is achieved when ATP-sensitized neutrophils are triggered with an allosteric FFAR2 modulator (i.e., Cmp58 or AZ1729, respectively). The signals generated by the orthosteric agonist propionate in neutrophils sensitized by the allosteric FFAR2 modulators AZ1729 or Cmp58 are balanced in the sense that the activation of the superoxide anion producing NADPH-oxidase is accompanied by a transient rise in [Ca²⁺]_i. It is clear from earlier data on the effects of the allosteric modulator AZ1729 being G protein biased that FFAR2 signaling can be mediated through both Gαi- and Gαq-containing G proteins.² In neutrophils, direct FFA2R signaling is insensitive to Gαq inhibition. This suggests that signaling bypasses Gαq and possibly engages Gαi, but this cannot be verified without a Gαi-selective inhibitor that is not available today.^12,85,112 It should be noted, however, that the ATP-induced cross-talk activation of AZ1729 allosterically modulated FFAR2 depends on a Gαq-containing G protein.⁵⁹ The precise signals generated downstream of the Gαq protein that activate the allosterically modulated FFAR2s is not known and can only be speculated on, but our earlier published data on a similar process, by which desensitized FPRs are reactivated by signals generated by other GPCRs (see below and^{46,58,113,114}), suggest that the cross-talk NADPH-activation signal is not the same as that leading to a direct NADPH-activation in naïve cells.

5.4. Interdependent Neutrophil Activation by Two Allosteric FFAR2 Modulators

The neutrophil priming characteristics of an allosteric FFAR2 modulator termed AZ1729¹⁰⁷ are the same as those for the allosteric modulator (Cmp58) described above. The similarities include the effects on the response induced by orthosteric FFAR2 agonists and on activation through receptor cross-talk with P2Y₂R.⁵⁹ More important is that the AZ1729 modulator turns Cmp58, the earlier described specific allosteric FFAR2 modulator, into an activating ligand. Thus, in cooperation, the two nonactivating allosteric modulators potently activate neutrophils, and they do this in a reciprocal manner. That is, Cmp58 turns also AZ1729 into a potent neutrophil-activating ligand (Figure 6C). Moreover, the signaling profile downstream of activated FFAR2 differs from that induced when an orthosteric agonist is part of the activation; signaling is biased in that no transient rise in intracellular Ca²⁺ is induced during the interdependent activation triggered by the two allosteric modulators. It is notable that the signals generated by FFAR2 that activate the neutrophil NADPH-oxidase are linked to an activation of the PLC–PIP₂–IP₃–Ca²⁺ pathway when triggered by an orthosteric agonist, and the signal bypasses this pathway when triggered by the allosteric modulators. It remains to be determined if the G protein is bypassed downstream of the biased signaling FFAR2 or if it can be activated without any PLC–IP₃ mediated rise in [Ca²⁺]_i. Moreover, the desensitized receptor is also functionally selective: the level of superoxide generated by the neutrophil NADPH-oxidase when triggered by the orthosteric agonist is lowered, whereas signaling through the PLC–PIP₂–IP₃–Ca²⁺ pathway is primed.⁵⁹ On the basis of the data presented and the fact that Cmp58 as well as AZ1729 (the two modulators) lack the carboxylic acid structure shown to be of vital importance for the ability of orthosteric agonists to activate FFAR2,¹¹⁵ it is reasonable to assume that the modulators interact with allosteric receptor sites and that that FFAR2 contains (at least) two such sites. The precise FFAR2 sites recognizing Cmp58 and AZ1729, respectively, are not known, but it is clear that one modulator turns the other into a potent neutrophil-activating agonist and this activity is inhibited by an orthosteric antagonist. This suggests that the two modulators not only interact with different allosteric binding sites but also that these sites at least partly overlap with the orthosteric binding site. In addition, the presented data support earlier findings showing that there is no direct link between the Ca²⁺ signaling route and the signals that activate the neutrophil NADPH-oxidase.^{21,58,113,116} Also, and possibly more importantly, the data presented show that two different conformational changes mediating FFAR2 activation are induced when the allosterically modulated receptor is activated by a second modulator and when the orthosteric binding site is occupied by a specific agonist, respectively. The precise signals that induced an assembly of the NADPH-oxidase components and an activation of the system are not known in detail. However, it is clear that the FFAR2 signaling response induced by an orthosteric agonist is balanced (activation of both the PLC–PIP₂–IP₃–Ca²⁺ and NADPH-oxidase routes), whereas that induced by two allosteric modulators is functionally selective with no involvement of the G protein linked activation of the PLC–PIP₂–IP₃–Ca²⁺ route.

6. Cross-Talk Signaling between Neutrophil GPCRs

6.1. Reactivation of Desensitized Receptors

The active signaling phase of agonist-occupied GPCRs is normally terminated through a receptor desensitization process. The basic mechanism for the transfer from a signaling to a nonsignaling state is the binding of β-arrestin or cytoskeletal proteins to parts of the receptor that are exposed to the cytosol.^22,39 The binding of a large protein physically obstructs interaction with other molecules and separates the receptor from the signaling heterotrimeric G protein and blocks binding of new G proteins; binding of a cytosolic blocking protein thus transfers the agonist-occupied receptor to a nonsignaling state. However, a novel form of the receptor cross-talk signaling process can reactivate desensitized receptors. This process has been extensively studied using neutrophils with desensitized FPRs that become reactivated by signals generated when the desensitized cells are activated by PAF (recognized by the PAFR) or ATP (recognized by P2Y₂R).^{12,58,113,114} The signals generated by the reactivated FPRs are biased in nature as these occur without any FPR-mediated rise in [Ca²⁺]_i. The precise signals generated by the PAFRs and the P2Y₂Rs that reactivate desensitized FPRs have not yet been identified, but the NADPH-oxidase activation signals generated by direct activation of FPRs and reactivation of the same receptors differ. This is illustrated by the effects of the phosphatase inhibitor calyculin A: this phosphatase inhibitor primes the response induced by direct activation of FPRs, whereas the receptor cross-talk reactivation is inhibited.⁵⁸ It is also clear that the reactivation signals triggered by the PAFRs and the P2Y₂R are generated downstream of the Gαq subunit present in the G protein that is activated by these receptors.¹²

Notably, the allosteric FFAR2 modulator Cmp58 has off-target effects on the neutrophil response induced by FPR1 and FPR2 agonists at concentrations too low to activate their respective receptor by themselves, a characteristic not shared by the orthosteric FFAR2 agonist propionate.¹¹² This suggests that not only P2Y₂R but also other receptors such as the Gαi-coupled FPRs have the capacity to activate the allosterically modulated FFAR2 from inside the plasma membrane, in a process very similar to the receptor cross-talk leading to a reactivation of desensitized FPRs.

6.2. Receptor Reactivation and the Role of β-Arrestin Recruitment

We have earlier shown that despite a lower potency of the non-β-arrestin recruiting F2Pal₁₀ in direct activation of the neutrophils this agonist acts as a full agonist in PAF-induced FPR2 reactivation.¹¹³ On the basis of this, we previously suggested that β-arrestin recruitment may play a critical role in receptor cross-talk-induced receptor reactivation. However, this suggestion has been challenged by our recent data obtained with phenol soluble modulating peptides (PSMs). These bacteria-derived formylated peptide toxins are encoded for and exported by community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) strains, and two of these (PSMα2 and PSMα3) are potent agonists recognized by FPR2. Similar to other FPR2 agonist, PSMα2 and PSMα3 activate neutrophils and induce a rise in the cytosolic concentration of free Ca²⁺ and a production and release of NADPH-oxidase derived reactive oxygen species (ROS). In contrast to conventional FPR2 agonistic peptides, the PSMα peptides lack the ability to promote β-arrestin recruitment.⁶⁶ Also, PSMα-desensitized neutrophils are reactivated through the receptor cross-talk mechanism, but the results obtained with the PSMα peptides clearly show that there is no direct link between β-arrestin recruitment and the potency of the cross-talk induced FPR2 reactivation.⁶⁶ Taken together, the available data thus strongly suggest that FPR2 desensitization and reactivation occurs primarily through the actin cytoskeleton and is independent of β-arrestin recruitment.

7. Future Challenges and Perspectives

It is clear that the development of new sophisticated cellular and molecular biology techniques during the last 10–20 years has made it possible to generate an impressive amount of new knowledge about basic signaling and the mechanisms that regulate the functional output downstream of activated/modulated GPCRs. There are however several more general GPCR challenges that are unmet, and there are also specific challenges concerning receptors expressed in innate immune cells such as the neutrophils. With respect to the prototype GPCR in neutrophils (FPR1) as well as to other neutrophil GPCRs, some important challenges are in part highlighted in earlier reviews dealing with the FPRs,^41,46 including the following: (i) the functions of GPCRs expressed in neutrophils of mammals other than Homo sapiens; (ii) the basis at a structural level, for the variety of selective agonists (and possibly antagonists/allosteric modulators) recognized by pattern recognition neutrophil GPCRs; and (iii) mechanisms utilized by neutrophil GPCRs in mediating both proinflammatory and resolving effects. Even if these questions are of immediate interest also today, new challenges have turned up in recent years. The progress in the identification and characterization of allosteric GPCR modulators (positive as well as negative) suggested to interact with a specific receptor either from outside or from inside the plasma membrane and the description of biased signaling ligands opens for possibilities far different from the conventional on/off receptor signaling scenario. Additionally, the finding that nonsignaling (desensitized receptors) can resume signaling trough a reactivation process, suggests that an additional layer of complexity will have to be added in models for GPCR activation. Particularly, the fact that cross-talk signaling between different GPCRs can trigger reactivation of desensitized receptors could be highly relevant not only for how neutrophils react to multiple gradients of different chemoattractants but also for how GPCRs can be activated in other cell types. The precise cross-talk signals generated by one receptor which activate another receptor are not yet known, and whether this type of signaling mechanism applies to other neutrophil GPCRs and to GPCRs expressed in other cells remains to be investigated. Studies in disease model systems, using mice that lack different Fprs, suggest that even if this group of receptors have critical roles in regulating the immune response and in host defense, the translation of data from animal models into human settings should be done with caution. It is clear that many agonists/antagonists shown to be potent activating/inhibiting ligands for human GPCR have low affinity (or are even inert) for their counterparts expressed in mouse cells. This strongly implies that potent/specific agonists/antagonists that are recognized by both murine and human receptors are needed for animal model studies and for future GPCR-based drug development.

Due to the increased knowledge during recent years about the defined receptor structures of several different GPCRs, this research field has entered a new era. When structural information for the neutrophil chemoattractant receptors (including both the FPRs and the FFARs) have been generated, this knowledge will make it possible to study their signaling and functional properties in more detail and offer conformational insights into activation induced both by conventional and biased signaling ligands when interacting with naïve receptors as well as with allosterically modulated receptors. This type of knowledge will possibly facilitate the elucidation of the mechanisms underlying not only the desensitization process but also of the reactivation of desensitized receptors. Our increased knowledge about the mechanisms that regulate neutrophil GPCR activities of prime importance during normal physiological conditions and in pathological settings will provide the means for the development of new prophylactic and secondary prevention strategies, as well as for effective treatments for diseases involving inflammatory or autoimmune processes. Signaling molecules that initiate inflammatory reactions as well as resolve an ongoing inflammation have been reported to mediate their effects through neutrophil GPCRs, but the precise mechanisms that regulate these two receptor reactions have not yet been defined. Studies in which signaling by ligands possessing initiating/pro-inflammatory or resolving/anti-inflammatory activities or of ligands that initiate/regulate both these activities will possibly provide new information about how the different functional and signaling pathway changes in relation to the structures of the ligands. We believe that neutrophil GPCR research during the coming years will generate data that substantially will increase our understanding about signaling and regulatory mechanism by which GPCRs, including the pattern-recognition receptors FPRs and FFARs, transduce the information from different inflammatory mediators to different biological functions.

The authors declare no competing financial interest.