Selective Localization of Recognition Complexes for Leukotriene B4 and Formyl-Met-Leu-Phe within Lipid Raft Microdomains of Human Polymorphonuclear Neutrophils¹

Lipid raft microdomains are generally envisioned as small dynamic regions within plasma membranes with characteristically high concentrations of sphingolipids and cholesterol. Lipid rafts, although heterogeneous, are enriched in many proteins engaged in neutrophil activation and migration, including TLR, chemokine receptors, multiple signaling molecules (Lyn and other src kinases, PI3K, Rho GTPases, G proteins, Src homology region 2 domain-containing phosphatase 1), and cytoskeleton-related proteins (1, 2, 3, 4, 5, 6, 7). The highly saturated sphingolipids generally pack tightly together in cholesterol-stabilized gel-like states termed liquid-ordered (Lo)³ arrays. This physical state is conducive to concentrating and constraining the mobility of multiple proteins to facilitate assembly of competent signaling complexes. Lipid rafts distribute uniformly throughout the plasma membranes of quiescent leukocytes but reorganize dramatically as cells polarize to form large aggregates at the leading edges and uropods, with aggregates at each location possessing highly distinct ganglioside and protein content (2, 8, 9, 10).

The dynamic polarization of lipid rafts and their capacity to regulate activation signaling and cytoskeletal rearrangement foster the hypothesis that lipid rafts are critically important to the mechanisms governing cellular locomotion and, in particular, directional migration. This may occur at many levels. First, lipid rafts may be the locus of chemotactic receptors that initiate cellular movement. CXCR1, CXCR4, and CCR5 chemokine receptors have been shown to localize in lipid rafts (2, 3, 11). Further, chemotaxin receptors characteristically couple to regulatory G proteins, and G proteins concentrate in lipid rafts (2, 3, 11). Lipid rafts may organize ion channels and other signaling elements within the plasma membrane to translate initial activation signaling into properly oriented signaling waves. We have shown that lipid raft disruption completely disables cells from propagating traveling Ca²⁺ waves along the plasma membrane of spontaneously polarized neutrophils (10). Not to the exclusion of important effects on upstream signaling, lipid rafts may also spatially couple the signaling apparatus to actin cytoskeleton and microtubules to achieve effective movement (3, 12). The observations that chemokine receptors and signaling intermediates localize within lipid rafts has raised the possibility that it may be a general property that chemotaxin recognition systems are fully functional only in lipid raft environments. Defining the extent to which chemotaxin recognition systems necessarily colocalize in lipid rafts would be of great value in understanding not only lipid raft function but also the potential for physical interplay between different chemotaxin recognition complexes within the same microdomains. In this study, we sought to determine whether receptors for two major nonchemokine neutrophil chemotaxins, leukotriene B4 (LTB4) and the bacterial peptide fMLP, partition to lipid rafts either constitutively or with stimulation and whether lipid raft integrity is necessary for activation signaling through these receptors.

Neutrophils were isolated using the method of Boyum (13) from peripheral blood obtained from healthy volunteers according to a protocol approved by the University of Michigan Institutional Review Board for Human Subject Research. Briefly, citrate-anticoagulated blood was sedimented with 6% dextran (0.9% NaCl), the erythrocytes were removed by hypotonic lysis, and neutrophils were isolated by density gradient centrifugation on a 10% Ficoll-Hypaque cushion. The resulting cells (>95% neutrophils) were washed in PBS and transferred to the appropriate buffer for further studies.

Fractionation of neutrophil membranes into lipid raft and nonraft fractions was performed using standard methods, with minor modifications (14, 15). Neutrophils were stimulated as described (∼2 × 10⁸ cells per preparation), and immediately collected by centrifugation in buffer A (25 mM HEPES, 150 mM NaCl, 1 mM EDTA (pH 7.0), with 1.4 μg/ml pepstatin A, 1.4 μg/ml leupeptin, 100 U/ml aprotinin, 4 mM iodoacetic acid, and 100 μM PMSF) at 4°C. The cells were then lysed in buffer A with 2.5 mM diisopropylfluorophosphate, 1 mM sodium orthovanadate, and 1.0% Brij-58 (Sigma-Aldrich) for 30 min at 4°C. The lysates were then layered on discontinuous sucrose gradients of 40-30-10%. After centrifugation at 180,000 × g for 20 h, the fractions were removed in 1-ml volumes. Protein concentrations were determined with a microBCA assay (Pierce). Samples of each fraction, adjusted to equal protein content, were examined by Western blotting as above, using the following primary Abs: rabbit anti-human B leukotriene receptor-1 (BLT-1) (120111; Cayman Chemical); and goat anti-human formyl peptide receptor (FPR) (sc-13193), rabbit anti-human flotillin-1 (sc-25506), and rabbit anti-human Cdc42 (sc-87X), all from Santa Cruz Biotechnology.

FITC-conjugated murine anti-human BLT-1 Ab (FAB099F) NAS obtained from R&D Systems. Anti-Cdc42 and anti-flotillin-1 Abs (Santa Cruz Biotechnology) were conjugated with FITC or tetramethylrhodamine isothiocyanate with standard kits (Invitrogen Life Technologies/Molecular Probes) and purified by gel chromatography. Fluorescein-conjugated formyl-Nle-Leu-Phe-Nle-Tyr-Lys (F1314) was obtained from Molecular Probes. Washed neutrophils were labeled with the appropriate Ab or fMLP analog, washed thoroughly, and fixed with 4% paraformaldehyde without being permeabilized. An Axiovert-135 inverted fluorescence microscope with HBO-100 mercury illumination (Zeiss) interfaced to a Dell 410 workstation via a Scion SG-7 video card (Vay Tek) was used. A narrow bandpass-discriminating filter set was used with excitation at 485DF20 nm and emission of 530DF30 nm for FITC. For tetramethylrhodamine isothiocyanate, an excitation of 540DF20 nm and an emission of 590DF30 nm were used (Omega Optical). Long pass dichroic mirrors at 510 and 560 nm were used for FITC and rhodamine, respectively. Single-cell spectra were obtained using an imaging spectrophotometer system. Labeled cells were illuminated with an excitation filter at 485DF22 nm and a 510LP dichroic mirror for resonance energy transfer experiments (16, 17). The emission spectra were obtained with an Acton-150 imaging spectrophotometer fiberoptically coupled to a microscope. The exit port of the spectrophotometer was attached to a Gen-II intensifier coupled with an I-MAX-512 camera (Princeton Instruments). Spectra collection was controlled by a high speed Princeton ST-133 interface and a Stanford Research Systems DG-535 delay gate generator and analyzed with Winspec software (Princeton Instruments).

After polymorphonuclear neutrophils were treated as indicated, lysates were prepared for Western blots according to a method adapted from the work of Gilbert et al. (18). Polymorphonuclear neutrophils were washed and resuspended in buffer A with protease inhibitors and mixed 1:1 (v/v) with boiling lysis buffer (62.5 mM Tris-HCl (pH 6.8) with 4% SDS, 8.5% glycerol, 5% 2-ME, 2 mM orthovanadate, 10 μg/ml aprotinin, 10 mM p-nitrophenylphosphate, and 0.025% bromphenol blue) and boiled for another 7 min. These lysates were then mixed 3:1 with four times sample buffer (0.25 M Tris (pH 6.8), 31% glycerol, and 8% SDS), boiled for 5 min, and electrophoresed on 8–16% gradient polyacrylamide gels, transferred to polyvinylidene difluoride membranes, labeled as indicated, and developed by ECL. Polyclonal goat anti-human Abs to total and phosphorylated ERK1/2, and HRP-conjugated donkey anti-goat secondary Ab, were obtained from Santa Cruz.

Cells were loaded (5 × 10⁶/ml) with the Ca²⁺-sensitive fluorescent dye fluo-3-acetoxymethyl ester (2 μM; Molecular Probes) at 30°C for 30 min in 145 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 10 mM glucose, 4 mM probenecid, and 10 mM HEPES, pH 7.4. After pretreatments as indicated, 2.5 × 10⁶ cells were suspended in 1 ml of incubation buffer and prewarmed to 37°C. Fluorescence intensities were then measured with a SLM8000 spectrofluorometer equipped with SLM Spectrum Processor v3.5 software (SLM Instruments), using a 1-cm light path cuvet at an excitation wavelength of 505 nm and an emission wavelength of 530 nm. Fluorescence intensities were acquired at 2-s intervals with continuous stirring of the cell suspension. These measurements were converted to nanomolar concentrations of [Ca²⁺]_i by the calibration method of Grynkiewicz et al. (19), using a K_d for fluo-3 of 864 nM (20).

Comparisons of means were performed with t tests using a p value of 0.05 to determine significance (GraphPad Prism version 3.00 for Windows; GraphPad).

Although every method for assessing whether specific proteins associate with lipid rafts is susceptible to artifact and misinterpretation, the most widely used of these has exploited two general properties of lipid rafts: their relatively low buoyant density; and their resistance to non-ionic detergents (14, 15, 21). Membrane fractionation experiments were performed to examine the association of high affinity LTB4 and fMLP receptors (BLT-1 and FPR, respectively) with lipid rafts, using a relatively nonselective detergent extraction protocol with 1.0% Brij-58. In unstimulated neutrophils (Fig. 1, top), virtually all the membrane BLT-1 is distributed in lower density lipid raft fractions where it coisolates with flotillin-1, a protein highly enriched in lipid rafts (7). By contrast, FPR partitions almost exclusively to the dense nonraft fractions where it codistributes with Cdc42, a protein largely excluded from lipid rafts in unstimulated neutrophils (22). In preliminary studies, we saw virtually identical distributions of BLT-1 and FPR when the cells were extracted with 0.5% Brij-58 (not shown). Similar results were also obtained when the cells were extracted with 1% Triton X-100, the detergent usually used for lipid raft fractionation, although in these experiments the yield of BLT-1 was compromised (not shown). This seemed to be a function of poor recovery rather than BLT-1 being partially soluble in 1% Triton X-100, given that BLT-1 did not appear in dense fractions with other soluble proteins. Experiments were next performed to determine whether cellular stimulation affects the partitioning of BLT-1 or FPR within membrane fractions. Neutrophils stimulated with LTB4 (5 × 10⁻⁸ M) for 20 s (corresponding to the timing of the peak in ERK 1/2 phosphorylation; see below) or for 5 min resembled unstimulated cells, because BLT-1 continued to partition to the lipid raft fractions whereas FPR remained in the nonraft fractions (Fig. 1, A and C, bottom panels). Likewise, stimulating neutrophils with fMLP for 90 s (corresponding to the timing of the peak in ERK 1/2 phosphorylation) or for 5 min did not change the partitioning of either BLT-1 or FPR (Fig. 1, B and D, bottom panels). To guarantee an adequate yield of lipid raft fractions for these experiments, all the neutrophils obtained from each donor were dedicated to a single stimulation protocol and membrane fractionation. For this reason, and also because the Western blots were not standardized between experiments, it is not possible to conclude from the results in Fig. 1 that stimulation with LTB4 or fMLP affected the total amounts of any of these proteins in the plasma membrane. Our findings do indicate, however, that the receptor/signaling complexes for fMLP and LTB4 are physically segregated from one another within neutrophil plasma membranes, both constitutively and after stimulation. Further, the disparity in the distributions of BLT-1 vs FPR is relatively robust with respect to the conditions of membrane fractionation. Still, it is possible that proteins could be selectively lost or redistributed during detergent extraction, so we sought corroboration by using FRET to examine the lipid raft localization of these proteins in intact cells.

FRET was performed to determine whether BLT-1 on unstimulated neutrophils could be found in association with flotillin-1. As shown in Fig. 2,A, there is substantial FRET with this labeling scheme, indicating molecular proximity (within a few nanometers) between flotillin-1 and BLT-1. In control experiments, there was no 590 nm emission in the presence of donor fluorochrome alone, and there was no FRET between flotillin-1 and Cdc42, the respective lipid raft and nonraft markers (not shown). By contrast, formyl-Nle-Leu-Phe-Nle-Tyr-Lys, used to detect FPR, produced robust FRET with Cdc42 (Fig. 2,D), but none with flotillin-1 (Fig. 2,C). This pattern agrees entirely with the membrane fractionation studies (Fig. 1), placing FPR predominantly in nonraft fractions. In preliminary experiments (not shown), we also evaluated GM1 and GM3 gangliosides as lipid raft markers and found very strong concordance with the results obtained with flotillin-1.

The association of BLT-1, but not FPR, with lipid raft markers suggests that lipid raft integrity may selectively influence the capacity for signal transduction through these receptors. To address this question, previously unstimulated human neutrophils were loaded with fluo-3 to measure [Ca²⁺]_i changes in response to chemotaxins. The cells were pretreated with methyl-β-cyclodextrin (Sigma-Aldrich) (MβCD; 5 mM for 5 min), a cyclic heptasaccharide that binds cholesterol avidly and extracts it from the plasma membrane, thereby destabilizing the Lo state of lipid rafts (22, 23). Preliminary experiments established that under these conditions, MβCD reduced plasma membrane cholesterol content by 17.4 ± 0.9% (p < 0.01), using a cholesterol oxidase assay (24). We have also shown that this protocol reduces the liquid ordering of neutrophil lipid raft aggregates (10). Following pretreatment, the cells were stimulated with fMLP (5 × 10⁻⁷ M) or LTB4 (5 × 10⁻⁸ M). Experiments confirmed that MβCD under these conditions did not adversely affect cell viability, as determined by trypan exclusion and by confirming a stable baseline [Ca²⁺]_i (data not shown). Also, flow cytometry confirmed that MβCD pretreatment did not significantly affect the levels of BLT-1 on the plasma membrane (not shown). As shown in Fig. 3, pretreatment with MβCD did not significantly affect the magnitude of the initial increase in [Ca²⁺]_i in response to fMLP, although the duration of the recovery phase was shortened. It has been demonstrated previously that MβCD has no direct effect on intracellular Ca²⁺ mobilization, but it does directly inhibit capacitative Ca²⁺ entry through the plasma membrane, thereby accounting for a hastened recovery of [Ca²⁺]_i even when the initial [Ca²⁺]_i flux is unperturbed (25). By contrast, pretreatment with MβCD under identical conditions substantially reduced initial Ca²⁺ mobilization in response to LTB4, in keeping with the localization of BLT-1 in lipid rafts. To exclude a nonspecific effect of cyclodextrins, cells were pretreated with α-cyclodextrin (α-CD; Sigma-Aldrich), which has poor affinity for cholesterol. α-CD (5 mM, 5 min), did not suppress the [Ca²⁺]_i response to LTB4.

It is possible that some of the effects of MβCD on intracellular Ca²⁺ mobilization could be considerably distanced from the initial engagement between BLT-1 and/or FPR with signaling kinases. Accordingly, we sought to assess the role of lipid raft integrity in another immediate signal transduction event. Fig. 4 shows that MβCD pretreatment completely abrogates the increased phosphorylation of ERK 1/2 tyrosine kinase in response to stimulation by 10⁻⁸ M LTB4. The effect was completely agonist specific, given that MβCD had no effect on fMLP (5 × 10⁻⁷ M)-induced ERK 1/2 phosphorylation. To confirm that the effect of MβCD is not a cyclodextrin-related artifact, we demonstrated that α-CD had no effect on ERK 1/2 phosphorylation. In addition, the effects of MβCD were duplicated with 2-β-hydroxypropylcyclodextrin (Sigma-Aldrich; 5 mM, 5 min), a cyclodextrin with cholesterol affinity that is comparable with that of MβCD.

All the preceding experiments have described the relationship between lipid rafts and activation signaling through fMLP and LTB4 in previously unstimulated neutrophils. It is well established that neutrophil activation signaling pathways are highly interactive, with some agonists capable of selectively desensitizing cells to subsequent stimulation by other agonists. The underlying mechanisms for this, however, are not fully defined. The following experiments were performed to determine whether the organization of lipid rafts influences the interaction between LTB4- and fMLP- mediated activation signaling. The ability of prior stimulation with LTB4 or fMLP to modulate subsequent responsiveness to the other agonist was examined, using initial intracellular Ca²⁺ mobilization as evidence of a signaling response (Fig. 5). As shown in Fig. 5,B, prior stimulation with fMLP completely negates responsiveness to LTB4 within 1 min (the recovery time for fMLP-induced intracellular Ca²⁺ transients). This desensitization lasts up to 10 min (data not shown). By contrast, prior stimulation with LTB4 had no effect on subsequent transients in response to fMLP stimulation (Fig. 5,A). The corresponding membrane fractionation experiments (Fig. 1) showed that neutrophils stimulated with LTB4 or fMLP did not change the partitioning characteristics of either BLT-1 or FPR. Collectively, these data demonstrate that 1) BLT-1 retains its physical separation from FPR within distinct microdomains in stimulated neutrophils and 2) heterologous desensitization of BLT-1 by FPR does not require that the two receptors be physically located within the same plasma membrane microdomains.

The Singer-Nicholson model of the plasma membrane describing a homogeneous protein-imbedded phospholipid bilayer has been replaced by a model incorporating heterogeneous lipid raft microdomains, enriched in sphingolipids and cholesterol and with distinctive and varied protein content (reviewed in Refs. 1 and 26, 27, 28). Many signaling proteins, including heterotrimeric G proteins and tyrosine kinases, concentrate at the cytofacial aspect of lipid rafts (28). For this reason, it has been proposed that lipid rafts provide the physical infrastructure for concentrating multiple elements to form signaling assemblies (29, 30). This engenders lateral organization among signaling cascades, potentially increasing their efficiency by keeping key elements in close proximity and by avoiding inappropriate cross-talk between pathways. Presently, there is relatively little information as to how lipid rafts regulate activation signaling in neutrophils.

Chemotaxin receptors, although diverse in their ligand specificity, share certain structural features such as seven membrane-spanning domains and binding to regulatory G proteins (i.e., G protein-coupled receptors) (11). It has been shown that the chemotaxin receptors CXCR1, CXCR4, and CCR5 localize to lipid rafts either constitutively or after activation by their ligands (2, 3, 11). Prior work has raised the possibility that most, if not all, chemotaxin recognition systems involving G protein-coupled receptors are located in lipid rafts and require lipid raft integrity to interact effectively with G proteins and downstream signaling elements. Our data show that this is not necessarily so. As shown in Figs. 3 and 4, fMLP is fully capable of activating at least some neutrophil signaling pathways (Ca²⁺ mobilization, ERK 1/2 phosphorylation) despite lipid raft disruption with MβCD. The existing literature regarding putative lipid raft associations of fMLP receptors illustrates the difficulties of establishing whether a protein is associated with lipid rafts. Our data agree with those of Pierini et al. (31) who showed that fMLP-induced up-regulation of β₂ integrins was not sensitive to MβCD. Our findings also agree with those of Barabé et al. (25), who found that MβCD did not affect the peak Ca²⁺ response to fMLP but did suppress the later sustained response due to a direct effect of MβCD on capacitative Ca²⁺ entry through the plasma membrane. Pierini et al. (31) found only partial suppression of the initial fMLP-induced increase in intracellular Ca²⁺ and a greater effect on the sustained response. Tuluc et al. (32) reported that MβCD could disrupt Ca²⁺ influx and ERK 1/2 phosphorylation in response to ≤10 nM fMLP, but the effect diminished with concentrations similar to those used in the present study. It is difficult to reconcile their findings with ours, partly because we did not examine suboptimal concentrations of fMLP, and Tuluc et al. pretreated neutrophils with MβCD under conditions exceeding those that we have found to be injurious to the cells. Also, in their study, lipid raft associations were only inferred from the effects of MβCD. Demonstrating cholesterol dependence alone is insufficient to prove that a signaling apparatus is located within lipid rafts. Cholesterol binds to many receptors and modulates their function, so it does not necessarily follow that effects of cholesterol extraction must be attributed to lipid raft destabilization (33). Moreover, cholesterol binding/extraction does not have entirely predictable effects on lipid raft structure, and the possibility of cholesterol binding agents acting nonspecifically often cannot be excluded (21, 28). As an example, Xue et al. (34) found that MβCD disrupted FPR-mediated Ca²⁺ signaling, but the presence of FPR in lipid rafts was inconclusive because a general colocalization with GM1 ganglioside could not be corroborated by colocalization with flotillin-1 or by FPR partitioning with lipid rafts in fractionated membranes. Also, this was demonstrated in FPR⁺ U937 cells, so their observations may be a function of cell type or signaling through a transfected FPR. Although it is possible that some signaling events initiated by FPR could be tied to lipid raft organization, our study shows that FPR functions independently of lipid rafts in engaging at least two downstream signaling pathways. Further, we also demonstrated the physical exclusion of FPR from lipid rafts, both by membrane fractionation and by FRET. FPR could be present in a previously described high density subset of lipid rafts (7), and our membrane fractionation scheme would not have recognized this, but this subset would have to lack flotillin-1 and ganglioside raft markers, and also be insensitive to MβCD, to conform with our results. By contrast, an identical experimental approach showed that BLT-1 is found virtually exclusively in lipid rafts. This was demonstrated by a complete concordance between its partitioning characteristics in membrane fractionation experiments, molecular coupling with lipid raft markers by FRET, and by the dependence of LTB4-mediated signaling on membrane cholesterol. The association of BLT-1 with lipid rafts was constitutive and persists after LTB4 or fMLP stimulation. To our knowledge, this is the first demonstration that BLT-1 is a lipid raft-associated protein. The segregation of BLT-1 from FPR within the plasma membrane is particularly striking, considering that many of the downstream signaling events and complex effector functions (i.e., chemotaxis) closely resemble those elicited by fMLP.

Neutrophils responding to a chemotaxin assume a polarized morphology with a distinct group of proteins, including receptors and downstream signaling intermediates, redistributing in or near the lamellipodium (reviewed in Ref. 3). It appears that lipid raft reorganization has much to do with maintaining this polarity. Some proteins redistribute within GM1 ganglioside-enriched lipid raft aggregates at the tail end of polarized cells within or near the uropod, while a distinct population of GM3 ganglioside-enriched lipid rafts, containing an equally distinct profile of proteins, aggregate in or near lamellipodia. The nonraft membrane of polarized cells predominates over the remaining cell body. This redistribution was demonstrated mostly by monitoring lipid raft-associated proteins and ganglioside markers, but we have also demonstrated the polarization of lipid rafts by directly assessing the degree of liquid ordering in the neutrophil plasma membrane (10). Lipid raft-associated proteins implicated in cell movement and found at the leading edges include GPI-linked proteins (CD87, CD59), and many signaling intermediates (2, 9, 22, 35, 36, 37, 38, 39). Lipid rafts at the rear contain CD44, CD43, PSGL-1, ICAM adhesion proteins, and ezrin-radixin-moesin actin-binding proteins (8, 9). This evidence certainly implicates lipid rafts in the formation of asymmetrical, spatially restricted signaling domains in migrating cells. The functional importance is further reinforced by evidence that lipid raft disruption suppresses traveling Ca²⁺ waves emanating from the lamellipodium, actin polymerization, Rac retention, uropod integrity, formation of lamellipodia, membrane ruffling, and chemotaxis (2, 9, 10, 31, 40, 41). In the present study, we examined neutrophils that were either unstimulated or exposed to uniform concentrations of agonists so we could examine the relevent signaling pathways in a system that does not involve cellular polarization. The distinctly separate partitioning of BLT-1 vs FPR persists through prolonged stimulation with LTB4 or fMLP (Fig. 1), indicating that the physical segregation of these receptors is remarkably stable compared with many others that are quite dynamic in their shifting associations with lipid rafts (28). A logical step for future work will be to determine whether FPR or BLT-1 distribute differently as neutrophils polarize toward a chemotactic gradient.

Our findings may also be relevent to the direction-finding aspect of chemotaxis. Fundamentally, chemotaxis is a computational problem. Cells must not only detect a chemotaxin, but also process this information to choose a direction of migration. It has been difficult to conceptualize how cells detect small chemotactic gradients in a noisy in vivo environment, but two mechanisms have been proposed. A spatial model is based on comparisons of chemotaxin concentrations between the lamellipodium and uropod, and a temporal model proposes that the cell periodically compares current chemotaxin concentrations with previous ones. It appears that fMLP may act through a temporal mechanism (42). It seems possible that lipid raft-associated chemotaxins may act by a spatial mechanism driven by rafts whereas fMLP may act by a temporal mechanism linked with non-raft membrane domains.

Another hypothetical consideration is that the spatial organization of chemotaxin recognition systems may affect the ways in which neutrophils respond to multiple signals. Chemotaxins induce transient responses in neutrophils, indicating that their recognition systems have incorporated mechanisms to terminate signaling, and also to desensitize cells to subsequent exposures to the same chemotaxin. Some chemotaxins such as fMLP can also desensitize neutrophils to other chemotaxins, such as IL-8, platelet activating factor, and LTB4 (43, 44, 45, 46). A consistent hierarchy of desensitization has been observed for a number of signaling events (43, 44, 45, 46). fMLP is the most effective at desensitizing neutrophils to other chemotaxins, whereas LTB4 ranks lower in the hierarchy by virtue of its inability to desensitize cells to fMLP. Our findings fit well with this formulation (Fig. 5). Many mechanisms, none mutually exclusive, have been described to explain heterologous receptor desensitization (46, 47, 48). Our results (Figs. 1 and 5) demonstrate that heterologous desensitization to BLT-1 by fMLP does not require colocalization within the same microdomains and that in fact the relevent signaling events can traverse the boundaries between lipid raft and nonraft microdomains.

The authors have no financial conflict of interest.