Chemotaxis and Calcium Responses of Phagocytes to Formyl Peptide Receptor Ligands Is Differentially Regulated by Cyclic ADP Ribose1

Cyclic ADP ribose (cADPR) is a calcium-mobilizing metabolite that regulates intracellular calcium release and extracellular calcium influx. Although the role of cADPR in modulating calcium mobilization has been extensively examined, its potential role in regulating immunologic responses is less well understood. We previously reported that cADPR, produced by the ADP-ribosyl cyclase, CD38, controls calcium influx and chemotaxis of murine neutrophils responding to fMLF, a peptide agonist for two chemoattractant receptor subtypes, formyl peptide receptor and formyl peptide receptor-like 1. In this study, we examine whether cADPR is required for chemotaxis of human monocytes and neutrophils to a diverse array of chemoattractants. We found that a cADPR antagonist and a CD38 substrate analogue inhibited the chemotaxis of human phagocytic cells to a number of formyl peptide receptor-like 1-specific ligands but had no effect on the chemotactic response of these cells to ligands selective for formyl peptide receptor. In addition, we show that the cADPR antagonist blocks the chemotaxis of human monocytes to CXCR4, CCR1, and CCR5 ligands. In all cases, we found that cADPR modulates intracellular free calcium levels in cells activated by chemokines that induce extracellular calcium influx in the apparent absence of significant intracellular calcium release. Thus, cADPR regulates calcium signaling of a discrete subset of chemoattractant receptors expressed by human leukocytes. Since many of the chemoattractant receptors regulated by cADPR bind to ligands that are associated with clinical pathology, cADPR and CD38 represent novel drug targets with potential application in chronic inflammatory and neurodegenerative disease.

A denosine 5Ј-diphosphate ribosyl cyclases (cyclases), such as the mammalian ecto-enzyme CD38, transform NAD ϩ into several products including the calcium-mobilizing metabolite cyclic ADP-ribose (cADPR) 3 (1,2). Since CD38 is a member of a highly conserved family of cyclases isolated from plants, invertebrates, and vertebrates (2), it has been hypothesized that CD38, via its production of cADPR, is likely to be an important regulator of calcium-based signal transduction. cADPR modulates the level of intracellular free calcium in cells in at least two ways. In combination with free cytosolic calcium, cADPR induces intracellular calcium release from ryanodine receptor-gated stores by a process referred to as calcium-induced calcium release (3). In addition, cADPR has been shown to regu-late the influx of extracellular calcium (4,5), possibly by activating the store-operated calcium release-activated calcium current channels (I crac ) (6). Although it is very clear that cADPR modulates intracellular free calcium levels in cells, less is known about which receptors rely on cADPR for signaling. In addition, very little is understood about the role(s) for cADPR in regulating important cellular processes such as development, growth, and differentiation.
To address which receptors utilize cADPR for signaling, several laboratories have now synthesized a number of different cADPR inhibitors including 8-Br-cADPR, a potent cADPR antagonist (7) and N(8-Br-A)D ϩ , a NAD ϩ analogue that can be cyclized by CD38 into the cADPR antagonist 8-Br-cADPR (5). These antagonists have recently been successfully used to identify receptors such as the muscarinic receptor that mobilizes calcium in a cADPR-dependent fashion (8,9). To assess the in vivo signaling role of cADPR, we produced mice that lack CD38 (10), one of the two known mammalian ADP-ribosyl cyclases (2). Using bone marrow neutrophils isolated from the CD38 knockout (KO) mice, we demonstrated that calcium signaling induced upon ligation of the classical chemoattractant formyl peptide receptor (FPR) is dependent on CD38 and cADPR (5). Importantly, we also found that chemotaxis of mouse neutrophils to the FPR ligand fMLF is regulated by cADPR and CD38 (5). Furthermore, we showed that pretreatment of normal mouse neutrophils with either 8-Br-cADPR or N(8-Br-A)D ϩ inhibited the chemotactic response of these normal neutrophils to fMLF (5,11). Together, these data showed that cADPR and the ADP-ribosyl cyclase CD38 modulate FPR-induced signal transduction and control the chemotactic responses of mouse neutrophils to fMLF.
The G protein-coupled FPR is one of the founding members of the chemoattractant receptor superfamily (12)(13)(14)(15). Like many of

Isolation of neutrophils and monocytes
Mouse bone marrow neutrophils were prepared by flushing bone marrow from tibias and femurs of mFPRI-deficient and C57BL/6J mice and then positively selecting the neutrophils using biotinylated GR-1 (BD PharMingen, San Diego, CA) and MACS Streptavidin Microbeads (Miltenyi Biotec, Auburn, CA). Purity was Ն95% as assessed by FACS. Human leukocytes were isolated from fresh peripheral blood donated by healthy volunteers in accordance with the Trudeau Institute Institutional Review Board regulations (samples kindly provided by the Blood Donor Center, Champlain Valley Plattsburgh Hospital, Plattsburgh, NY). Neutrophils were purified (Ն95% purity) using a one-step Ficoll gradient (Robbins Scientific, Sunnyvale, CA). Monocytes were isolated by enriching for mononuclear cells using the one-step Ficoll density gradient centrifugation method and then purified (Ն95%) by MACS using a CD14 monocyte isolation kit (Miltenyi Biotec). All purified cells were washed and resuspended in HBSS supplemented with 1% FBS.

Analysis of CD38 expression and cyclase activity in human peripheral blood leukocytes
Human peripheral blood leukocytes were isolated from whole fresh peripheral blood by Ficoll density gradient centrifugation and then assessed for CD38 expression by FACS. Cell suspensions were stained with mouse anti-human CD38-biotin (Caltag Laboratories, Burlingame, CA) or biotinylated mouse isotype control Ab (Zymed Laboratories, San Francisco, CA) and anti-human CD15-FITC (BD Biosciences, San Jose, CA) and then streptavidin-allophycocyanin (BD Biosciences), and were then analyzed by flow cytometry using a FACSCalibur (BD Biosciences).
To measure CD38-dependent ADP ribosyl cyclase activity in neutrophil lysates, purified human peripheral blood neutrophils (1.5 ϫ 10 7 cells) were disrupted with lysis buffer containing 2 mM EDTA, 1 mM DTT, 2 g/ml leupeptin, 1 g/ml pepstatin, 50 g/ml PMSF, and 1% Triton X-100 (v/v; Sigma-Aldrich). Solubilized proteins were recovered from the lysate by centrifugation and were incubated with biotinylated mouse anti-human CD38 (Caltag Laboratories) or a biotinylated mouse IgG1 Ab (Zymed Laboratories) along with streptavidin agarose beads (Sigma-Aldrich). The precipitated protein-bead complexes were extensively washed and then resuspended in 40 l of HBSS. CD38-dependent GDP-ribosyl cyclase activity was determined by measuring accumulation of the fluorescent product, cyclic GDP-ribose (cGDPR), as previously described (41). Aliquots of the protein-bead complex (10 l) were placed in individual wells of opaque 96-well plates (Corning, Rochester, NY) containing 80 l of HBSS/well and were allowed to settle for 30 min. NGD ϩ (10 l, 40 M final concentration) was added to each well and the plates were incubated for 20 min at 37°C. Cyclase activity was then determined by monitoring the accumulation of cGDPR in the reaction using a SpectraMax GeminiXS microplate fluorometer (Molecular Devices, Sunnyvale, CA) that was calibrated for an excitation wavelength of 300 nm and an emission wavelength of 415 nm. Relative fluorescence units are reported.

Chemotaxis assays
Chemotaxis assays with mouse neutrophils and human neutrophils and monocytes were performed using 24-well Transwell plates (Costar, Cambridge, MA) with a 3-m (for neutrophil chemotaxis assays) or a 5-m (for monocyte chemotaxis assays) pore size polycarbonate filter. Chemotaxis assays for N9 cells were performed with 48-well chemotaxis chambers (NeuroProbe, Cabin John, MD) using polycarbonate filters with an 8-m pore size. Chemoattractants were diluted in HBSS and placed in the lower chamber, upper chamber, or both upper and lower chambers of the Transwell. In most experiments, cells were first pretreated for 15-20 min with either 8-Br-cADPR (0 -100 M; Sigma-Aldrich) or N(8-Br-A)D ϩ (500 M) and then added to the upper chamber of the Transwell in the continued presence of the drug. For neutrophil assays, 1 ϫ 10 6 cells/Transwell were incubated at 37°C for 45 min, whereas for monocyte and microglial cell assays 1 ϫ 10 5 cells/Transwell were incubated at 37°C for 90 min. The transmigrated cells were collected from the lower chamber, fixed, and counted on a flow cytometer. To determine the absolute number of cells in each sample, a standard number of 20 m size fluorescent microspheres (Polysciences, Warrington, PA) was added to each tube and counted along with the cells. The total number of transmigrated cells ϭ the number of counted neutrophils ϫ the total number of beads/the number of beads counted. In some cases the results are expressed as the mean Ϯ SD of the chemotaxis index (CI). The CI represents the fold increase in the number of untreated or inhibitor-pretreated cells that migrated in response to the chemoattractant divided by the basal migration of untreated or antagonist pretreated cells migrating in response to control medium.

Intracellular calcium mobilization
Purified human peripheral blood neutrophils or monocytes were resuspended in cell-loading medium (HBSS plus 1 mM Ca 2ϩ plus 1 mM Mg 2ϩ plus 1% FBS plus 4 mM probenecid) at 1 ϫ 10 7 cells/ml and loaded with a mixture of the calcium-sensitive dyes Fluo-3 AM (4 g/ml) and Fura-Red AM (10 g/ml; Molecular Probes, Eugene, OR). The cells were incubated at 37°C for 30 min, washed twice, and resuspended in cell-loading medium at 1 ϫ 10 6 cells/ml. Cells were preincubated in the presence or absence of 8-Br-cADPR (100 M; Sigma-Aldrich) for 20 min and then stimulated with various chemokines and chemoattractants in calcium-containing or calcium-depleted (plus 2 mM EGTA) medium. The accumulation of intracellular free calcium was assessed by FACS over the next 7 min by measuring the fluorescence emission of Fluo-3 in the FL-1 channel and Fura-Red in the FL-3 channel. Data were analyzed by FACS analysis software FlowJo 4.0 (Tree Star, San Carlos, CA) using the kinetic platform. The relative intracellular free calcium levels were expressed as the ratio between Fluo-3 and Fura-Red mean fluorescence intensity over time.

Statistical analysis
Data sets were analyzed using GraphPad Prism version 3.0a for Macintosh (GraphPad Software, San Diego, CA). Student's t test analyses were applied to the data sets to determine statistically significant differences between groups. Differences were considered significant when p values were Յ0.05.

Chemotaxis of mouse bone marrow neutrophils to mFPR1 and mFPR2 agonists is regulated by cADPR
Murine neutrophils express at least two functional fMLF receptors (mFPR1 and mFPR2) that can be activated by a number of different agonists, including fMLF, HIV-derived peptides, SAA polypeptide, and amyloid ␤ peptide (14,15). We previously showed that 8-Br-cADPR, a competitive antagonist of cADPR that acts by blocking cADPR-induced calcium mobilization (7), also inhibits chemotaxis of mouse bone marrow neutrophils to fMLF (5,43). To test whether 8-Br-cADPR inhibits the migration of mouse neutrophils to other mFPR1 or mFPR2 ligands, we measured the chemotactic response of 8-Br-cADPR-treated neutrophils isolated from the bone marrow of C57BL/6 and mFPR1deficient mice (Ref. 23; mFPR1 KO) to peptides that activate mFPR1 and/or mFPR2. In agreement with previous studies (44), we found that HIV-derived T20 peptide induced a strong migratory response in C57BL/6J neutrophils and essentially no response in mFPR1 KO neutrophils (Fig. 1), indicating that T20 peptide preferentially activates mFPR1. In contrast, fMLF, A5 peptide, and the HIV-derived F peptide induced the chemotaxis of mFPR1 KO neutrophils (Fig. 1B), indicating that these ligands can activate mFPR2 in mouse bone marrow neutrophils. However, all peptides including fMLF, A5 peptide, and F peptide induced even greater migration of C57BL/6J neutrophils compared with mFPR1 KO neutrophils (cf Fig. 1, A vs B), suggesting that these ligands, at least at the concentrations used in this experiment, may activate both mFPR1 and mFPR2 in mouse bone marrow neutrophils. Interestingly, 8-Br-cADPR treatment blocked the migration of both C57BL/6J and mFPR1 KO neutrophils to all of the mFPR1 and mFPR2 ligands tested (Fig. 1). This 8-Br-cADPR-mediated inhibition of chemotaxis to mFPR1 and mFPR2 ligands was specific for FPR ligands as 8-Br-cADPR treatment had no effect on the chemotaxis of mouse neutrophils to the CXCR2 agonists IL-8 and MIP-2 (data not shown and Ref. 5). Together, these data show that a cADPR antagonist can be used to inhibit the migration of mouse bone marrow neutrophils to a variety of FPR ligands and suggest that cADPR regulates the signaling of both mFPR1 and mFPR2 in mouse neutrophils.

Human neutrophils express a functional ADP-ribosyl cyclase
The cADPR antagonist 8-Br-cADPR inhibited the chemotaxis of mouse neutrophils to a number of different FPR ligands (Fig. 1). These data suggested the intriguing possibility that compounds that block cADPR-dependent signaling could be used therapeutically to modulate inflammatory responses mediated by the potentially pathogenic FPRL1 ligands. Since CD38 is the primary and best-characterized mammalian ADP-ribosyl cyclase (cyclase) (11,45,46), we first performed FACS analysis to determine whether CD38 is expressed by different subpopulations of human peripheral blood leukocytes. As has been previously reported (47), only a subset of the CD15 neg lymphocytes expressed CD38 while essentially all of the CD15 low monocytes expressed CD38 ( Fig. 2A). Although it has been previously suggested that CD38 is not expressed by human neutrophils (47), we found that the majority of the highly granular CD15 high -expressing neutrophils expressed CD38, albeit at lower levels compared with the CD38-expressing monocytes and lymphocytes ( Fig. 2A).
Primary human lymphocytes and monocytes have been previously shown to exhibit CD38-dependent ADP-ribosyl cyclase activity (48,49); however, it has not been tested whether primary human neutrophils express a functional cyclase. To test whether the cyclase reaction could be catalyzed by neutrophil-derived CD38, we immunoprecipitated CD38 from lysates of purified peripheral blood neutrophils and then incubated the purified CD38 protein with a synthetic substrate, nicotinamide guanine dinucleotide (NGD ϩ ). Cyclases, such as CD38, utilize NGD ϩ as a substrate and catalyze production of the highly fluorescent cyclic product cGDPR, which can be detected using a fluorometer (41). As shown in Fig. 2B, CD38 isolated from human neutrophils rapidly produced cGDPR when incubated with NGD ϩ . Similar results were obtained upon analysis of polymorphonuclear cell (PMN) samples from multiple donors (data not shown) indicating that human neutrophils, like human lymphocytes and monocytes, express CD38 and can produce cyclic metabolites such as cADPR.

The cADPR antagonist 8-Br-cADPR blocks the chemotaxis of human neutrophils to FPRL1-specific ligands
Human neutrophils express at least two FPRs, the high-affinity fMLF-binding receptor FPR and the low-affinity fMLF-binding receptor FPRL1 (15). The human FPRL1 is considered to be an orthologue of the mFPR2 and both have a similar low affinity for fMLF (25, 50 -52). Likewise, the human FPR is thought to be most closely related to the mFPR1, although the affinity of the mFPR1 for fMLF is 100-to 500-fold lower than the affinity of the human FPR for fMLF (25, 50 -52). Using cells transfected with either FPR or FPRL1, it has been demonstrated that some agonists, such as fMLF and T20 peptide, specifically activate FPR while others, like A5 and amyloid ␤ peptide, preferentially activate FPRL1 (reviewed in Refs. 14 and 15). Since the cADPR antagonist 8-Br-cADPR blocked the chemotaxis of mouse neutrophils to all of the FPR ligands that we tested, including agonists that are known to specifically activate human FPR or FPRL1, we predicted that 8-Br-cADPR would also inhibit the chemotaxis of human neutrophils to all of the different FPR-and FPRL1-binding ligands. To test this hypothesis, we incubated human peripheral blood neutrophils in the presence or absence of 8-Br-cADPR and then measured the migration of these cells to fMLF and to the synthetic FPRL1specific agonist A5 peptide (28). Interestingly, treatment of human neutrophils with 8-Br-cADPR had no effect on the migration of these cells to nanomolar concentrations of fMLF (Fig. 3A). However, the cADPR antagonist did inhibit the migration of human neutrophils to the A5 peptide (Fig. 3B) in a dose-dependent man-ner (Fig. 3C). Thus, the cADPR antagonist blocked chemotaxis of human PMNs to a FPRL1-specific ligand but not to a FPR-specific ligand.
Since the cADPR antagonist inhibited neutrophil migration to the A5 peptide, a FPRL1-specific ligand, but had no effect on chemotaxis to very low concentrations of fMLF, we postulated that the cADPR antagonist inhibited FPRL1 dependent, but not FPR FIGURE 2. Human neutrophils and monocytes express CD38, an enzyme with ADP-ribosyl cyclase activity. A, CD38 expression levels were determined on human peripheral blood leukocytes by FACS using fluorochrome-conjugated anti-human CD38 (or isotype control Ab) and antihuman CD15. Cell size (forward scatter, FSC) and granularity (side scatter, SSC) of the peripheral blood leukocytes was analyzed to identify lymphocytes (Lym), monocytes (MN) and PMNs. The expression of CD38 was then assessed in PMN (CD15 high ), monocyte (CD15 low ), and lymphocyte (CD15 neg ) cells. The percentage of cells in each quadrant is indicated. B, ADP-ribosyl cyclase activity in lysates from human peripheral blood neutrophils was measured using a fluorometric assay. CD38 was immunoprecipitated from protein lysates prepared from human peripheral blood neutrophils and was then incubated in medium (f) or in the presence of the CD38 substrate, NGD ϩ (F). Accumulation of the fluorescent cyclic product, cGDPR, was measured over time and is expressed as relative fluorescence units (RFU). The data are representative of at least 10 similar experiments using neutrophils from different donors. dependent, signaling in human neutrophils. To test this hypothesis, we incubated human neutrophils in the presence or absence of 8-Br-cADPR and then performed migration assays using IL-8 and a variety of different chemoattractants with demonstrated specificity for either FPR or FPRL1. As shown in Fig. 4A and Table I, we found that 8-Br-cADPR treatment of human neutrophils had no effect on migration of the cells to IL-8. However, unlike what we had observed with murine neutrophils (Fig. 1), 8-Br-cADPR treat-ment of human neutrophils had minimal effect on the chemotactic response of the cells to fMLF or T20 peptide ( Fig. 4A and Table  I). Strikingly, however, the migration of human neutrophils to all of the FPRL1 ligands tested, including HIV-derived F peptide, HIV-derived V3 peptide, amyloid ␤ peptide, MMK-1 peptide, and A5 peptide, was significantly inhibited by 8-Br-cADPR treatment ( Fig. 4A and Table I). Taken together, these data indicate that the cADPR antagonist 8-Br-cADPR specifically inhibits the migration  of human neutrophils to FPRL1-specific, but not FPR-specific, ligands.

cADPR regulates calcium mobilization in FPRL1-stimulated human neutrophils
Since cADPR is a known calcium-mobilizing metabolite (2), we predicted that cADPR might regulate intracellular free calcium levels in FPRL1-stimulated human neutrophils. To test this hypothesis, we measured the accumulation of intracellular free calcium in primary human neutrophils stimulated with IL-8, FPRspecific agonists, or FPRL1-specific agonists (Fig. 4B). As expected, calcium mobilization in response to IL-8 was essentially identical between 8-Br-cADPR-treated and control neutrophils (Fig. 4B). Likewise, 8-Br-cADPR treatment had minimal effect on the calcium response of human neutrophils stimulated with the FPR-specific ligands fMLF and T20 peptide (Fig. 4B). In contrast, the calcium response of 8-Br-cADPR-treated neutrophils to the FPRL1-specific ligands, A5 and MMK-1 peptides, was reduced compared with the neutrophils that were not pretreated with 8-Br-cADPR (Fig. 4B). Interestingly, calcium mobilization in FPRL1stimulated neutrophils was completely dependent on the presence of external calcium because the calcium signal was abolished when external calcium was depleted from the medium using EGTA (Fig. 4B). This is in contrast to what we observed when the cells were stimulated with IL-8, T20 peptide, or fMLF, as a substantial calcium response was seen even when extracellular calcium was depleted from the medium (Fig. 4B). Thus, these data indicate that FPR and FPRL1 induce distinct calcium responses, with FPR inducing a calcium response that occurs independently of cADPR and is largely, although not exclusively, due to intracellular calcium release, while FPRL1 induces a calcium response that is regulated by cADPR and is primarily due to extracellular calcium influx.

cADPR antagonists block FPRL1-mediated chemotaxis, but not FPRL1-induced chemokinesis
Treatment of human neutrophils with the cADPR antagonist did not completely block migration of the neutrophils to the FPRL1 ligands (e.g., Fig. 4A). However, we have previously shown that 8-Br-cADPR specifically blocked the chemotaxis (chemoattractant-induced directional migration) of murine bone marrow neutrophils to fMLF, but had minimal effect on fMLF-induced chemokinesis (chemoattractant-induced nondirectional migration) (5).
Therefore, we concluded that cADPR-dependent signaling is most critical for mediating directional movement toward FPR ligands.
To determine whether cADPR is required for the directional movement of FPRL1-stimulated human neutrophils, we incubated peripheral blood neutrophils in the presence or absence of 8-Br-cADPR and then performed a Transwell checkerboard assay to measure basal migration, chemokinesis, and chemotaxis. As shown in Fig. 5, A and B, pretreatment of human neutrophils with 8-Br-cADPR had no effect on basal migration, chemokinesis, or chemotaxis of neutrophils stimulated with the FPR ligands, T20 peptide, and fMLF. Likewise, treatment of neutrophils with 8-Br-cADPR had no effect on the basal migration or migration due to chemokinesis of neutrophils activated with the FPRL1 ligands, V3 peptide, F peptide, amyloid ␤ peptide, and A5 peptide (Fig. 5, C-F). In striking contrast, 8-Br-cADPR treatment inhibited the directional migration of the neutrophils responding to the FPRL1 ligands and reduced the migration that can be specifically attributed to chemotaxis by ϳ90% (Fig. 5, C-F). Therefore, these data demonstrate that the cADPR antagonist specifically blocks the directional movement of human neutrophils that are responding to FPRL1-specific signals.

The NAD ϩ analogue N(8-Br-A)D ϩ inhibits chemotaxis of FPRL1-activated neutrophils
Our experiments demonstrate that treatment of human neutrophils with a cADPR antagonist blocks the directional migration of these neutrophils to a number of FPRL1-specific ligands including inflammatory mediators such as amyloid ␤ peptide. These data strongly suggest that compounds that either inhibit the activity of the cyclase(s) that produce cADPR or that alter the products that are produced by these cyclase(s) could also be used to block the migration of neutrophils responding to FPRL1-specific agonists. One easily synthesized compound that can be used to alter product formation by ADP-ribosyl cyclases is the NAD ϩ analogue N(8-Br-A)D ϩ (40). ADP-ribosyl cyclases, like CD38, utilize this NAD ϩ analogue as a substrate, but instead of producing cADPR, cells incubated with this substrate produce 8-Br-cADPR, the cADPR antagonist (7). In previous experiments using mouse neutrophils, we showed that treatment of neutrophils with N(8-Br-A)D ϩ inhibited the chemotactic response of the neutrophils to fMLF in a CD38-and cADPR-dependent manner (5, 11).
Since human neutrophils express CD38 and are dependent on cADPR for FPRL1-induced chemotactic responses, we predicted a Human peripheral blood neutrophils were preincubated for 15 min in medium alone or 100 M 8-Br-cADPR and then assessed in Transwell migration assays. b The chemoattractants (or media control) were placed in the bottom well of the Transwell chambers at the concentrations listed in Fig. 4  that treatment of human neutrophils with N(8-Br-A)D ϩ would inhibit the capacity of these cells to respond to chemotactic gradients of FPRL1-specific agonists. To test this prediction, we preincubated human peripheral blood neutrophils with N(8-Br-A)D ϩ and then assessed the chemotactic potential of these cells in checkerboard assays using FPR and FPRL1 agonists as the chemoattractants. As shown in Fig. 6A, treatment of neutrophils with N(8-Br-A)D ϩ had absolutely no effect on the chemokinetic or chemotactic  response of the cells to the FPR ligand fMLF. In addition, the NAD ϩ analogue did not affect the chemokinetic response of the neutrophils to the FPRL1 ligands amyloid ␤ peptide and HIVderived V3 peptide (Fig. 6, B and C). However, the chemotactic response of neutrophils responding to FPRL1 ligands was dramatically reduced when the cells were preincubated with N(8-Br-A)D ϩ (Fig. 6, B and C). Together, these results show that both NAD ϩ analogues and cADPR antagonists can be used to block the directional migration of human neutrophils to inflammatory chemoattractants that activate FPRL1 but not FPR.

8-Br-cADPR inhibits FPRL1/mFPR2-mediated chemotaxis of multiple cell types
Neutrophils are not the only cells that express FPRL1 and migrate directionally in response to FPRL1 ligands. Indeed, FPRL1 is constitutively expressed by monocytes (15) and is inducibly expressed by microglial cells (39), which reside in the CNS and are of myeloid origin. Since FPRL1-dependent migration is regulated by cADPR in neutrophils, we postulated that the cADPR antagonist 8-Br-cADPR could also be used to block the migration of other cell types to FPRL1 ligands. To examine this possibility, we tested the effect of 8-Br-cADPR treatment on migration of human monocytes in response to FPRL1-specific ligands. First, we assessed the chemotactic response of 8-Br-cADPR-treated human peripheral blood monocytes to FPR and FPRL1-specific ligands. As previously reported, peripheral blood monocytes migrated in response to a variety of FPR and FPRL1 ligands including HIV-derived peptides (Fig. 7A), amyloid peptides (Fig. 7B), and synthetic peptides (Fig. 7, A and B). Treatment of monocytes with 8-Br-cADPR had no effect on the chemotactic responses of these cells to the FPR ligands, fMLF and HIV-derived T20 peptide (Fig. 7, A and  B). In contrast, the migration of the 8-Br-cADPR-treated monocytes was significantly reduced in response to the synthetic FPRL1 ligands MMK-1 and A5 peptides, the HIV-derived F peptide, and the amyloid peptides SAA and amyloid ␤ peptide (Fig. 7, A and B). Therefore, in complete accordance with our data using peripheral blood neutrophils, we found that the cADPR antagonist inhibited the chemotaxis of primary human peripheral blood monocytes to a number of different FPRL1-specific ligands. Next, we examined whether cADPR controls the migration of mouse myeloid-derived microglial cells to mFPR1/mFPR2 ligands. To do so, we used a murine microglial cell line, N9, that expresses typical markers of resting mouse microglia and are frequently used for functional analyses of microglial cells (18,39,53). Importantly, when N9 cells or normal mouse microglial cells are stimulated with LPS the ADP-ribosyl cyclase CD38 is expressed at low levels on the plasma membrane (data not shown), and mFPR2, the mouse orthologue of human FPRL1, is also upregulated (18,39). As expected, after 24 h of LPS treatment, the N9 cells expressed mFPR2 and could migrate in response to several mFPR2 ligands including the synthetic W and A5 peptides and high micromolar concentrations of fMLF (Fig. 7C). However, when the N9 cells were first pretreated with the cADPR antagonist 8-Br-cADPR, the cells were significantly impaired in their ability to migrate in response to any of the agonists (Fig. 7C). Interestingly, the cADPR antagonist had no effect on the chemotaxis of the microglial cells to another chemoattractant, C5a (Fig. 7C). Taken together, these data indicate that the cADPR antagonist specifically blocks FPRL1/mFPR2-dependent chemotaxis in myeloid-derived monocytes and microglial cells.

Treatment of monocytes with a cADPR antagonist blocks calcium influx and chemotaxis to homeostatic and inflammatory chemokines
The data from the experiments with human leukocytes indicated that cADPR is required for FPRL1-dependent chemotaxis but is not needed for CXCR1/2 (IL-8 receptors), FPR, or C5a receptordependent chemotaxis. These data suggested that cADPR must regulate signaling through only a subset of the chemoattractant receptors expressed by monocytes. Human peripheral blood monocytes are reported to express a number of additional chemokine receptors including CCR1, CCR2, CCR5, and CXCR4 (31,54). To test whether cADPR regulates signaling through some of these other chemokine receptors, we measured calcium and chemotactic responses in 8-Br-cADPR-treated human monocytes that were stimulated with CCR1/CCR5 and CXCR4 agonists (Fig. 8). As expected, the chemotactic response of 8-Br-cADPR-treated monocytes stimulated with the FPR ligand fMLF was completely normal (Fig. 8A). Likewise, the fMLF-induced calcium response was unaffected by the presence of 8-Br-cADPR (Fig. 8B). In contrast, 8-Br-cADPR treatment did block the migration of monocytes in response to the CCR1/CCR5 ligands RANTES and MIP-1␣ and to the CXCR4 ligand stromal cell-derived factor 1 (SDF-1) (Fig. 8A). Interestingly, migration of CCR1-expressing murine neutrophils to MIP-1␣ was also found to be cADPR and CD38 dependent (data not shown). Although 8-Br-cADPR treatment did not completely inhibit calcium mobilization in SDF-1-or MIP-1␣-stimulated monocytes, the response was significantly reduced (Fig. 8B). When the cells were stimulated in calcium-free medium, no calcium mobilization was observed after SDF-1 and Mip-1␣ stimulation (Fig. 8B). Thus, as we observed after stimulation with FPRL1-specific ligands (see Fig. 4B), calcium mobilization induced by ligation of CCR1/CCR5 and CXCR4 is almost exclusively due to extracellular calcium influx. Taken together, cADPR appears to modulate calcium influx in cells that have been activated by chemoattractant receptors that induce calcium mobilization via a calcium influx-dependent mechanism. This cADPR-regulated calcium influx signal is required for FPRL1-, CCR1/CCR5-, and CXCR4-induced chemotaxis but is not obligate for FPR-, CXCR1/2-, or C5aR-mediated chemotaxis.

Discussion
The data presented in this manuscript demonstrate that cADPR regulates calcium signaling in a discreet subset of chemokine and chemoattractant receptors. The chemokine receptors that utilized cADPR for signaling could be distinguished from the other non-cADPR-dependent receptors because the receptors that utilized cADPR mobilized calcium in a unique fashion. Typically, it is believed that engagement of chemokine receptors leads to the generation of the calcium-mobilizing second messenger inositol trisphosphate (IP 3 ) which in turn leads to release of intracellular calcium from IP 3 receptor-gated stores in the endoplasmic reticulum (55). The initial, very rapid, shortlived release of calcium is often, although not always, accompanied by a delayed and sustained influx of extracellular calcium (55). This classical calcium response to chemoattractant receptor engagement was observed after engagement of receptors such as C5aR (data not shown), CXCR1/2 (Fig. 4), and the human high-affinity FPR (Fig. 4); receptors that all signal in a cADPR-independent fashion. In striking contrast and in agreement with what we previously observed with mFPR receptors (5), engagement of the chemoattractant receptors that do utilize cADPR, such as FPRL1, CXCR4, CCR1, and CCR5, resulted in minimal intracellular calcium release that was accompanied by a very strong influx of extracellular calcium (Figs. 4 and 8).
Although it is not known how cADPR regulates extracellular calcium influx in any cell type, it has been suggested that cADPR regulates the activation of L-type calcium channels and store-operated I crac channels (6,56). Interestingly, human neutrophils have been reported to express L-type channels as well as I crac channels (57,58). Regardless of the exact mechanism by which cADPR regulates calcium influx in leukocytes, the data are clear that a specific cADPR antagonist partially inhibits the calcium response and blocks chemotaxis of human leukocytes responding to FPRL1, CXCR4, and CCR1/CCR5 ligands (i.e., Fig. 4).
These cADPR-dependent chemoattractant receptors are not the first chemoattractant receptors to be identified that mobilize calcium primarily by a calcium influx-dependent mechanism. Indeed, ligation of the platelet-activating factor receptor, CCR2, and CX 3 CR1 induces minimal intracellular calcium release and strong calcium influx (59 -62). Interestingly, no one has yet identified the calcium-mobilizing second messenger that potentiates calcium influx upon ligation of these receptors although it has been shown that IP 3 is apparently not involved in signaling through CCR2 (60). Therefore, based on these data we predict that other chemoattractant receptors such as platelet-activating factor receptor, CCR2, and CX 3 CR1 that primarily mobilize calcium via a calcium influxdependent mechanism may also utilize cADPR for signaling.
Chelating extracellular calcium with EGTA (Figs. 4 and 8) or using calcium-free medium (data not shown) appeared to completely abolish the calcium response of all the chemoattractant receptors that require cADPR for optimal calcium mobilization. However, the cADPR antagonist only partially inhibited calcium influx in response to FPRL1, CCR1/CCR5, and CXCR4 agonists (Figs. 4 and 8). These data indicate that other calcium-mobilizing second messengers, such as sphingosine-1-phosphate or perhaps IP 3 , must also play a role in modulating calcium influx in these cells (63,64). In addition, the data indicate that at least some chemoattractant-induced signals are likely to induce functional changes in the neutrophils even in the absence of cADPR. In fact, the neutrophils treated with the cADPR antagonist or the NAD ϩ analogue were activated by FPRL1 ligands and were perfectly competent to migrate nondirectionally (Figs. 5 and 6), indicating that activation-induced nondirectional movement is not controlled by cADPR but instead may be regulated by other calcium-mobilizing metabolites. Taken together, the data indicate that cADPR/ CD38 antagonists are not simply global inhibitors of calcium mobilization in chemoattractant-activated leukocytes and that these antagonists modulate the extracellular calcium influx induced by a selective group of chemoattractants and regulate only a subset of the biologic functions induced upon binding of these chemoattractants.
Our observation that two highly related human chemoattractant receptors, FPR and FPRL1, can be distinguished by their reliance on cADPR for calcium signaling is of potential therapeutic importance. FPRL1-dependent calcium mobilization is primarily due to extracellular calcium influx and utilizes cADPR, whereas FPRdependent calcium mobilization is due in large part to intracellular calcium release and is cADPR independent. This result was initially somewhat surprising to us since we had previously shown that cADPR and CD38 are necessary for chemotaxis of mouse neutrophils to the prototypic FPR receptor ligand fMLF (5). Indeed, the mFPR1 and human FPR are believed to be orthologues of one another and were presumed to signal through similar mechanisms (14,15). However, the data presented here, as well as previous published data, indicate that there are significant differences between these two receptors. First, the affinity of fMLF for human FPR is quite high (nanomolar range) whereas the affinity of fMLF for the mFPR1 is ϳ400 times lower (51). Second, mFPR1 expressed by mouse bone marrow neutrophils appears to be a more promiscuous receptor as it can be activated by high concentrations of ligands such as HIV-derived F peptide and the synthetic A5 peptide (Fig. 1), which do not efficiently activate the high-affinity human FPR (15,28). Third, chemotaxis induced upon mFPR1 engagement is inhibited by the cADPR antagonist (Fig. 1), whereas chemotaxis induced upon ligation of the human FPR is not affected by the cADPR antagonist (Fig. 3). Finally, fMLF only weakly induces extracellular calcium influx in human peripheral blood neutrophils (Fig. 4), while it strongly induces calcium entry in fMLF-stimulated mouse bone marrow neutrophils (5). Thus, while mFPR1 is most homologous at a structural level to human FPR (52), its affinity for fMLF, its reliance on cADPR, and the pattern of calcium mobilization induced by receptor engagement appears more similar to that seen with the human FPRL1 and mFPR2 receptors.
Despite the biochemical and functional differences between mFPR1 and human FPR, it is clear from the data presented here that a cADPR antagonist as well as a CD38 substrate analogue can be used to inhibit the chemotactic response of human and mouse neutrophils to a diverse array of mFPR2 and human FPRL1-specific ligands including several synthetic peptides, multiple HIV gp120-derived peptides, and at least two different amyloidogenic peptides (Figs. 4 and 6). This result was not limited to neutrophils as we also found that chemotaxis of human monocytes and mouse myeloid-derived microglial cells to these same ligands was inhibited by the cADPR antagonist (Fig. 7). Therefore, based on these data, we suggest that the signal transduction pathways engaged upon mFPR2/FPRL1 ligation in all three cell types are similarly dependent on cADPR. Furthermore, we predict that cADPR antagonists should be effective inhibitors of FPRL1-dependent chemotaxis of all myeloid-derived human cells. Finally, given that a number of the FPRL1-specific ligands are present in diseased tissue and are postulated to exacerbate the inflammatory response within these tissues (38), we predict that compounds like 8-Br-cADPR and N(8Br-A)D ϩ are likely to be useful for assessing the impact of FPRL1-induced inflammatory cell recruitment on the progression of pathology within the diseased tissues.