Functional selective FPR1 signaling in favor of an activation of the neutrophil superoxide generating NOX2 complex

Abstract The formyl peptide receptors FPR1 and FPR2 are abundantly expressed by neutrophils, in which they regulate proinflammatory tissue recruitment of inflammatory cells, the production of reactive oxygen species (ROS), and resolution of inflammatory reactions. The unique dual functionality of the FPRs makes them attractive targets to develop FPR‐based therapeutics as novel anti‐inflammatory treatments. The small compound RE‐04‐001 has earlier been identified as an inducer of ROS in differentiated HL60 cells but the precise target and the mechanism of action of the compound was has until now not been elucidated. In this study, we reveal that RE‐04‐001 specifically targets and activates FPR1, and the concentrations needed to activate the neutrophil NADPH‐oxidase was very low (EC50 ∼1 nM). RE‐04‐001 was also found to be a neutrophil chemoattractant, but when compared to the prototype FPR1 agonist N‐formyl‐Met‐Leu‐Phe (fMLF), the concentrations required were comparably high, suggesting that signaling downstream of the RE‐04‐001‐activated‐FPR1 is functionally selective. In addition, the RE‐04‐001‐induced response was strongly biased toward the PLC‐PIP2‐Ca2+ pathway and ERK1/2 activation but away from β‐arrestin recruitment. Compared to the peptide agonist fMLF, RE‐04‐001 is more resistant to inactivation by the MPO‐H2O2‐halide system. In summary, this study describes RE‐04‐001 as a novel small molecule agonist specific for FPR1, which displays a biased signaling profile that leads to a functional selective activating of human neutrophils. RE‐04‐001 is, therefore, a useful tool, not only for further mechanistic studies of the regulatory role of FPR1 in inflammation in vitro and in vivo, but also for developing FPR1‐specific drug therapeutics.

the form of formylated peptides, but also numerous nonformylated peptides/proteins/lipopeptides and other molecules such as small compounds and peptidomimetics. 3,8,9 FPR1 and FPR2 exhibit a large overall amino acid sequence similarity with a high degree of identity in the cytosolic parts and a lower degree in the extracellular domains. 4 This suggests that the two receptors differ more when it comes to ligand binding than in the intracellular signals transmitted. Nevertheless, many agonists cross activate the two receptors although there are a few reported that are highly specific for one or the other of the two receptors. 9 The downstream signals generated by agonists of FPRs regulate neutrophil directional migration (chemotaxis), mobilization of adhesion molecules to the cell surface, and secretion of inflammatory mediators, including proteolytically active proteases. Another feature of FPR agonists is also the activation of the electron transporting NADPH-oxidase complex type 2 (the NOX2 complex) with the capacity to produce superoxide anions (O 2 − ) that secondarily generates other reactive oxygen species (ROS). 10 Activation of neutrophils is essential for defense against microbes and for clearance of harmful tissue debris, but also to limit further neutrophil recruitment and facilitate tissue repair. Thus, these dual functions need to be tightly controlled through the different phases of inflammation. The effect of ROS shows a similar type of complex role when affecting different types of inflammation. ROS released in high quantities from neutrophils is generally regarded as driving acute inflammation. However, it is also clear that ROS could dampen inflammation, an effect more likely to operate in the resolution phases. [11][12][13] Hence, in light of this complex and so far not completely understood role of FPR and ROS regulation, our accumulated research proposes a regulatory role of ROS produced by the NADPH-oxidase in many cellular processes. 14,15 Patients, as well as experimental animals, with chronic granulomatous disease, lacking the ability to generate ROS, suffer not only from severe microbial infections, but also from a variety of inflammatory complications indicative of important functions of ROS in the mechanisms that control inflammation. [16][17][18] The importance of ROS in the regulation of inflammation also gains support from earlier studies in which we through positional cloning of a disease-linked genetic polymorphism, have identified Ncf1 (encoding for the p47 phox subunit of the NADPH-oxidase complex) as a diseaseassociated gene 19 and the molecular basis being linked to a compromised ROS production. 20 Similarly, polymorphism of Ncf1 plays a role in human autoimmune diseases, 21,22 and it has been shown to be of importance for disease severity of arthritis, psoriasis, colitis, and lupus in animal models (reviewed in Holmdahl et al. and Urbonaviciute et al. 14,23 ). Hence, it is apparent from both pharmacologic and genetic deletion studies, that FPRs have multiple roles in diseases conditions associated with a dysregulated inflammation. Mice deficient in individual FPRs show not only an increased susceptibility to microbial infections but also a delayed tissue repair. 7,24,25 In addition, a recent study has elegantly demonstrated that activation of FPRs improves cardiac function in a post myocardial infarction model, 26 suggesting an antiinflammatory/pro-resolving role of FPR agonists.
The introduction of the biased GPCR signaling concept rapidly became the starting point not only for more detailed characterization of known GPCR agonists but also for the search for new biased GPCR agonists that could be used to develop drug candidates. 27,28 The concept of biased signaling or functional selectivity describes that different ligands for a given receptor can stabilize receptor in different conformations allowing distinct signaling pathways with different functional activities. The concept has been shown to be valid also for FPR2 as illustrated by the downstream signaling by FPR2-specific agonistic lipopeptides/pepducins, peptidomimetics, as well as by formylated peptides generated by virulent Staphylococcus aureus bacteria. [29][30][31] These biased FPR2 agonists are potent triggers of a rise in intracellular calcium ([Ca 2+ ] i ) and a release of superoxide through the NADPH-oxidase but in contrast to earlier described FPR2 agonists, they lack the ability to recruit -arrestin and induce chemotaxis. [29][30][31] Due to the similarities between FPR1 and FPR2 it is reasonable to assume that FPR1 also can be stabilized in a conformation that opens for one signaling pathway downstream of the receptor but not for another. This assumption gains support from a study showing that selective formylpeptide analogues can discriminate between different biologic responses, being able to trigger chemotaxis but not activate the superoxide generating neutrophil NADPH-oxidase. 32 In attempt to identify novel ROS activators, we have earlier

Chemicals and reagents
The compound RE-04-001, with structure related to the class of compounds known as quinolones that were described in the patent application WO 2012/127214 and reported earlier in screening studies. 33 For intellectual property reasons, the chemical structure of RE-04-001 is not disclosed. For more information about RE-04-001 and to make it possible to reproduce the data presented herein, the compound will be provided to other researchers under a material transfer agreement

Isolation of human neutrophils and culture of neutrophil-like HL-60 cells
Neutrophil granulocytes were isolated from peripheral blood or buffy coats obtained from healthy adults. 35

Calcium mobilization
Neutrophils at a density of 5 × 10 7 cells/ml in KRG without Ca 2+ sup-

Neutrophil NADPH-oxidase activity
Neutrophil O 2 − production was determined using an isoluminolenhanced chemiluminescence (CL) system (details are given in Dahlgren et al. 37

Phosphorylation of ERK1/2 determined by electrochemiluminescence
Human neutrophils (2 × 10 6 /ml) were stimulated with fMLF or RE-04-001 for 2 min followed by rapidly cool down to stop reaction with ice old lysis buffer provided by Meso Scale Diagnostics (MSD, Rockville, MD, USA) according to manufacturer's instructions as described. 31

Data analysis
Data analysis was performed using GraphPad Prism 8.0 (Graphpad Software, San Diego, CA, USA). Curve fitting was performed by nonlinear regression using the sigmoidal dose-response equation (variable slope). Statistical analysis was performed on raw data values using either a repeated measurement 1-way ANOVA followed by Dunnett's multiple comparison post-hoc test or a paired Student's t-test. Statistically significant differences are indicated by *P < 0.05 and **P < 0.01.

RE-04-001 activates human neutrophils
A compound library containing drug-like small molecules was used in a screening study to identify novel NADPH-oxidase activators. 33 The release of O 2 − from neutrophil-like HL60 cells was determined, and RE-04-001 was found to activate the NADPH-oxidase (Fig. 1). In addition as secondary pharmacology screen, the activity of this compound (≥50% inhibition or stimulation) was determined using the HitProfil- Berridge 40 ). The results obtained with these antagonists clearly show that RE-04-001 is recognized by FPR1 (Fig. 2C). Based on the lack of

F I G U R E 1 The novel small compound RE-04-001 (for short-RE-in figures and legends) activates neutrophil-like HL60 cells. (A)
A sensitive technique to measure superoxide production was used to determine the ability of RE to activate neutrophil-like HL60 cells (10 5 cells). Cells were pre-incubated at 37 • C for 5 min before agonist stimulation (indicated by arrows) with RE (100 nM, solid line), the formyl peptide receptor ( activates neutrophils to release O 2 − , and there was a very rapid onset of the response that was then terminated in around 5 min after the initiation, a response pattern very similar to that induced by the two prototype FPR peptide agonists (Fig. 3A). The maximal level of O 2 − production induced by RE-04-001 was of the same magnitude as that induced by 100 nM fMLF, suggesting that RE-04-001 is a full agonist (Fig. 3A). The response induced by RE-04-001 was concentration dependent with an EC 50 value in the low nanomolar range ( Fig. 3B), which is much lower than that for the prototype FPR1 agonist fMLF (EC 50 ≈ 20 nM; Fig. 3B). In line with the data obtained with FPR-specific antagonists in the [Ca 2+ ] i assay system (Fig. 2C), the inhibitory profile for RE-04-001 was the same as that that of fMLF (sensitive to cyclosporine H but not to PBP 10 ) but different from WKYMVM (Fig. 3C).
The preference of RE-04-001 for FPR1 over FPR2 in human neutrophils gained further support from receptor homologous desensitization experiments. Neutrophils first activated with RE-04-001 were not only homologously desensitized (nonresponsive) to a second stimulation with RE-04-001 but were also refractory to stimulation with fMLF ( Fig. 3D). In contrast, these RE-04-001 desensitized cells were still fully responsive to a second stimulation with the FPR2 agonist WKYMVM (Fig. 3D). Taken together, these data show that RE-04-001 is a very potent stimulus that activates the neutrophil NADPHoxidase and this activation is achieved through signals specifically generated by FPR1.
It is well known that the NADPH-oxidase activity triggered by FPR-specific agonists is substantially increased in TNF primed neutrophils. 34,44 Accordingly, the amount of O 2 − produced by TNF primed cells was substantially increased with RE-04-001 as the activating FPR agonist (Fig. 3E). The increase due to priming with TNF was of the same magnitude as that with fMLF and WKYMVM (Fig. 3E inset). Finally, in agreement with the lack of inhibitory effect of the G q inhibitor on the FPR-mediated rise in [Ca 2+ ] i (Fig. 2D) NADPH-oxidase activity is shown for comparison (Fig. 3F). Taken together, these data show that RE-04-001 is a potent and full agonist selective for FPR1, and the agonist activates the neutrophil NADPHoxidase independent of coupling to a G q containing G protein.

Comparison of neutrophil chemotaxis induced by fMLF and RE-04-001
Based on the fact that the prototype FPR1 agonist fMLF and a large number of other earlier described FPR agonists potently recruit neutrophils, the FPRs are termed chemoattractant receptors. 3

RE-04-001 activates ERK1/2 phosphorylation rather than promoting -arrestin recruitment
The functional selectivity profile of RE-04-001 in human neutrophils suggests that the agonist triggers a biased signal cascade downstream FPR1. In addition to a rise in [Ca 2+ ] i , many FPR agonists trigger also ERK1/2 phosphorylation and recruitment of cytosolic -arrestin to cytoplasmic parts of the activated receptors. 45 For many GPCRs the latter event is of importance for receptor desensitization and internalization as well as for the transduction of noncanonical signals of which activation of ERK1/2 may be one. 46 Phosphorylation of ERK1/2 in human neutrophils upon agonist stimulation was determined as previously described. 31 Similar to potent agonistic activity by RE-04-001 in inducing a rise in [Ca 2+ ] i , the agonist induced also ERK1/2 phosphorylation and the potency was comparable to, or slightly higher than that of fMLF (Fig. 5A). The ability of RE-04-001 to promote receptor-mediated recruitment of -arrestin was studied in CHO cells overexpressing FPR1. 30 In contrast to the potent activity of RE-04-001 in inducing a transient rise in [Ca 2+ ] i and ERK1/2 phosphorylation ( Fig. 2A, 5A), the amount of -arrestin recruited by RE-04-001 in FPR1 overexpressing cells was negligible in comparison to that induced by fMLF (Fig. 5B). In agreement with the receptor specificity of RE-04-001, this agonist did not recruit any -arrestin in FPR2 overexpressing cells; the FPR2 agonist WKYMVM was included as an FPR2 control for comparison ( Fig. 5B inset).  to a homologous desensitized state in which the cells are nonresponsive to second agonist dose (Fig. 3D). There is a known hierarchy between different neutrophil GPCRs, and in this hierarchy FPR1 is ranked higher than the receptors for the cytokine IL-8 regarding both the NADPH-oxidase activation and neutrophil chemotaxis. 47,48 In accordance with this, the homologous FPR1 desensitization induced by fMLF and RE-04-001 was accompanied by a concomitant inhibition (heterologous desensitization) of the response mediated by the IL-8 receptors, making these FPR1 desensitized cells nonresponsive to IL-8 stimulation (Fig. 6A).

RE-04-001 promotes FPR1 to crosstalk with other neutrophil receptors
Recent research suggests that the receptor crosstalk hierarchy is complex and not only desensitized receptors but also allosteric modulated GPCRs can communicate with other receptors. 45 inhibited not only to an FPR1 antagonist but also by an antagonist specific for FFAR2 (Fig. 6B) shows that this response is achieved through receptor crosstalk between FPR1 and FFAR2.
Opposite to the heterologous inhibitory effect of RE-04-001 on the IL-8 response (Fig. 6A), a substantially enhanced PAF response no difference was observed in cells when desensitized by fMLF or RE-04-001 (Fig. 6C inset). The involvement of FPR1 in this response is evident from the fact that the second PAF response in RE-04-001 desensitized cells is sensitive to the FPR1 antagonist cyclosporine H when added just prior to PAF stimulation (Fig. 6D). This is in line with the earlier data showing that PAF/PAFR is able to transduce a not yet known signal leading to a reactivation of neutrophils with desensitized FPRs. 53 In summary, we show that the novel FPR1 agonist RE-04-001, despite its biased signaling feature, similar to fMLF places FPR1 in the same position in the neutrophil receptor hierarchy and allows receptor crosstalk with other GPCRs to either suppress or amplify the neutrophil response.

The termination of the RE-04-001-induced activation of the NADPH-oxidase is regulated primarily by the actin cytoskeleton rather than by -arrestin
Despite the fact that -arrestin plays an important role in receptor desensitization for many GPCRs, we and others have demonstrated that the actin cytoskeleton, rather than the recruited -arrestin, constitutes the basis for FPR desensitization and termination of signals that activate the ROS-producing oxidase. 45,54,55 This notion gains further support from the fact that FPR1 is homologously desensitized also by the non--arrestin recruiting agonist RE-04-001. In addition, in neutrophils pretreated with the actin cytoskeleton disrupting agent latrunculin A, RE-04-001-induced activation resulted in a 4-fold higher superoxide production in comparison to that produced by the cell in the absence of latrunculin A (Fig. 7A). Furthermore, RE-04-001 activated neutrophils transferred to a nonsignaling desensitized state were resensitized/reactivated and produce ROS when the actin cytoskeleton was disrupted through the addition of latrunculin A (Fig. 7B). These data, obtained with RE-04-001 as an activating agonist, are in agreement with the pattern induced by fMLF ( Fig. 7A and B). Taken together, we show that RE-04-001-induced FPR1 desensitization in neutrophils occurs primarily through the involvement of an intact actin cytoskeleton.

RE-04-001 is resistant to oxidation by the MPO-H 2 O 2 -halide system
Processing of NADPH-oxidase-derived hydrogen peroxide (H 2 O 2 ) by MPO, a neutrophil enzyme stored in the azurophil granules, results in a generation of highly reactive oxidants that regulate many biologic processes in addition to bacterial killing. 15 (Fig. 7C), the peptides earlier shown to be ROS sensitive. 34,57 No inhibition was seen when Cmp43 and Act-389949 were exposed to the MPO-H 2 O 2 -halide generating system (Fig. 7) results also in agreement with earlier findings. 34,58 Compared to the FPR1 selective peptide agonist fMLF, RE-04-001 was fairly resistant to the MPO-H 2 O 2 -halide radical system (Fig. 7C). Taken together, these data show that the RE-04-001 resists inactivation induced by the MPO-H 2 O 2 -halide system.

DISCUSSION
In this study, we show that the small molecule RE-04-001 activates human neutrophils and the agonist is specifically recognized by FPR1, one of the pattern-recognition FPRs. In-depth characterization of the has been reported to exert anti-inflammatory effects and protect mice from myocardial infarction injury, 59 whereas another compound (Bristol Meyers-Squibb; BMS-986235) has proceeded into a clinical phase I study as a selective FPR2 agonist for prevention of heart failure. 60 Yet another FPR2 selective agonist Act-389949 entered a clinical phase I study but the data obtained show that the surface exposed neutrophil receptors were rapidly lost, although the mechanisms for this were not described. 61  Thus, RE-04-001 clearly reveals a functional selective neutrophil response, and this was linked to a low level of -arrestin recruitment.
Our finding that the ability of FPR1 agonists to trigger neutrophil migration is linked to the ability to recruit -arrestin is in agreement with the emerging concept of biased FPR agonism and functional selectivity shown to be valid also for FPR2. 29 45 We have previously demonstrated that the activities of several other neutrophil GPCRs also are regulated by the actin cytoskeleton. 31,75,76 The biased signaling concept is now firmly established in GPCR biology. 27,28 Clearly, this concept is valid also for FPR1; in contrast to the prototype FPR1 agonist, RE-04-001 has a biased signaling profile.
Similar to RE-04-001, several FPR2 agonists have earlier been shown to transduce a biased signaling feature in neutrophils. [29][30][31] It is interesting to note that similar to RE-04-001, the biased signaling FPR2 agonists that lack ability to recruit -arrestin and are also poor neutrophil chemoattractants, 29-31 suggesting a role for -arrestin in regulating both FPR1-and FPR2-mediated directional cell migration. The nonpeptide compound termed Quin-C1 77 has also been shown to be a biased signaling FPR2 agonist, but with the reversed functional selectivity; it lacks the ability to trigger superoxide release, while being able to induce neutrophil chemotaxis. 77 The precise mechanism that deter-   82 ). In addition, downstream molecular mechanisms may differ between human, and mouse FPRs.
A prominent example is that the potent human FPR1 agonist fMLF is a very poor agonist for its mouse orthologue, 82