Binding of ligands to N-formyl peptide chemoattractant receptors exposed on human neutrophils generates signals in the cells that induce an activation of the superoxide anion producing NADPH-oxidase. Ligand binding is followed by a rapid association of the ligand-receptor complex with the cytoskeleton, a process leading to desensitization of the cells with respect to NADPH-oxidase activation. We show that neutrophils that have experienced an intracellular calcium rise obtained through interaction with the calcium-specific ionophore ionomycin are “primed” with respect to the FMLP-induced production of superoxide anions. Mobilization of FMLP receptors from intracellular pools is one well-known mechanism behind the primed response. Based on our finding that ionomycin-treated neutrophils could not be desensitized, we suggest that the lack of association between the ligand-receptor complex and the cytoskeleton is an additional priming mechanism. Since in vivo-exudated neutrophils, which also had mobilized intracellular organelles, could be desensitized, we suggest that the abolished desensitization in ionomycin-treated neutrophils is not due to an inability of newly recruited receptors to couple to the cytoskeleton. We show that a rapid termination of FMLP-induced superoxide anion production is obtained in both desensitizable and nondesensitizable neutrophils, suggesting that the desensitization phenomenon is of limited importance in the oxidase termination process.
The neutrophil-mediated defense against microbial infections is dependent on a unique ability of the phagocytes to produce large amounts of reactive oxygen species, i.e., superoxide anion and hydrogen peroxide (1, 2). These oxygen radicals are generated by an enzyme system, the NADPH-oxidase, which is assembled upon activation of the cells (3, 4). The NADPH-oxidase is triggered by many different stimuli, among which the chemotactic peptide FMLP is a prominent example (5, 6). Most neutrophil receptors (including the receptors for FMLP) are stored in subcellular organelles (for review, see 7 , and in order to respond properly to various stimuli, the neutrophils have to recruit the different intracellular granule subsets to the plasma membrane, mobilizing new receptors to the cell surface (8). This process also endows the neutrophil plasma membrane with increasing amounts of the membrane-spanning component (the b cytochrome) of the NADPH-oxidase (7, 9).
Binding of FMLP to the neutrophil FMLP receptor (FMLP-R) activates several signal transduction pathways (5, 10, 11). However, the precise signal(s) responsible for the subsequent activation of the NADPH-oxidase has not yet been identified. The amount of superoxide anion and hydrogen peroxide released upon neutrophil activation with FMLP is influenced not only by the degree of receptor exposure on the cell surface (12, 13) but also by the termination of the oxidase activity. The signals that lead to termination have not yet been defined. It has been suggested that the ability of the receptor-ligand complex to generate transmembrane signals is lost when the complex associates with the cytoskeleton, an event known to follow shortly after the binding of FMLP to its receptor. It has been convincingly shown that this cytoskeletal binding leads to desensitization of the receptor with respect to further activation by the agonist (14, 15, 16, 17). It has also been suggested that the coupling of the receptor-ligand complex to the cytoskeleton is responsible for the termination of an ongoing FMLP-induced NADPH-oxidase response (14, 15, 16).
It is well known that neutrophils that have mobilized their intracellular FMLP-R stores are “primed” with respect to FMLP-induced NADPH-oxidase activation. One mechanism behind this state must be an increased generation of second messengers that activate the NADPH-oxidase, but other regulatory mechanisms might also be of importance. The present study was performed to reveal whether priming following a rise in intracellular calcium is associated with a change in cytoskeleton-mediated regulation of FMLP receptor activity. The fact that desensitization was abolished in calcium ionophore-primed cells also made it possible to investigate the role of desensitization for the termination of the oxidative response.
Materials and Methods
Dextran and Ficoll-Pacque were purchased from Pharmacia (Uppsala, Sweden). FMLP, cytochalasin B, isoluminol, and Triton X-100 were obtained from Sigma (St. Louis, MO). Ionomycin was purchased from Calbiochem (La Jolla, CA). The phycoerythrin-conjugated mAb against CR3 was purchased from Becton Dickinson (San Jose, CA). The radiolabeled peptide [3H]FMLP was obtained from New England Nuclear (Boston, MA), and FITC-labeled formyl-Nle-Leu-Phe-Nle-Tyr-Lys (FITC-FNLPNTL)3 was from Molecular Probes (Eugene, OR).
Isolation of phagocytic cells
Blood neutrophils were isolated from heparinized whole blood or from buffy coats obtained from apparently healthy adults. After dextran sedimentation at 1 × g, hypotonic lysis of the remaining erythrocytes, and centrifugation in a Ficoll-Paque gradient (18), the neutrophils were washed twice and resuspended in Krebs-Ringer phosphate buffer containing glucose (10 mM), Ca2+ (1 mM), and Mg2+ (1.5 mM) (KRG, pH 7.3).
Exudated neutrophils were harvested from skin chambers placed on unroofed skin blister lesions on the volar surface of the forearms of healthy human volunteers, as previously described (19, 20). In each experiment, two chambers with three 0.6-ml wells covering the lesions were used. The chambers were filled with autologous serum, and the neutrophils were allowed to accumulate in the chambers for 24 h. More than 95% of the cells harvested from the chambers were neutrophils.
Measurement of neutrophil superoxide anion production
Neutrophil production of superoxide anion was assayed with an isoluminol-enhanced chemiluminescence (CL) system (21). The CL activity was measured in a six-channel Biolumat LB 9505 (Berthold Co., Wildbad, Germany), using disposable 4-ml polypropylene tubes with a reaction mixture of 0.9 ml containing 1 × 106 neutrophils, HRP (4 U), and isoluminol (a membrane-impermeable CL substrate, 2 × 10−5 M). This set-up measures the release of superoxide anion from the cells (21). The tubes were equilibrated for 5 min at 37°C, after which either the stimulus or, in some experiments, the cells were added. The light emission was recorded continuously. By a direct comparison of the SOD-inhibitable reduction of cytochrome c (using a millimolar extinction coefficient of 21.1 for cytochrome c (23, 24)) and SOD-inhibitable isoluminol-amplified CL, 7.3 × 107 counts were found to equal the production of 1 nmol of superoxide anion.
Mobilization of subcellular organelles
Neutrophil subcellular organelles were mobilized by treating the cells with ionomycin (22), a calcium-specific ionophore. After preincubating the cells for 5 min at 37°C, ionomycin (5 × 10−7 M final concentration) was added, and the incubation was continued for 5 min. The cells were then centrifuged, washed once, resuspended in KRG, and put on ice until used.
Neutrophils (107/ml) were incubated for 5 min at 15°C, FMLP (10−7 M final concentration) was added, and the incubation was continued for 10 min. The cells were then added (50–100 μl) to prewarmed (37°C) CL vials containing isoluminol, HRP, and FMLP (10−7 M final concentration), and the production of superoxide anion was followed as described above.
Determinations of receptor exposure by FACS analysis
Neutrophils were fixed for 30 min in ice-cold paraformaldehyde (4% w/v in PBS).
To determine the exposure of CR3, 10 μl of a conjugated mAb was added to a cell pellet (∼100 μl) of 106 cells. The cells were incubated on ice for 30 min, washed twice with KRG, and analyzed for amount of cell bound probe (correlating to the amount of CR3) by flow cytometry (FACScan; Becton Dickinson, Mountain View, CA).
To determine the exposure of FMLP-R, a FITC-conjugated formylated peptide (FITC-FNLPNTL; 10−8 M final concentration) was added to a cell pellet (∼100 μl) of 106 paraformaldehyde-fixed cells in the absence or presence of an excess amount (5 × 10−6 M) of nonlabeled FMLP. The cells were incubated at 22°C for 30 min, and no washing was performed after labeling. The amount of specifically bound probe (correlating to the amount of FMLP-R) was determined by flow cytometry (23).
The cellular content of F-actin
The F-actin content in the neutrophils was analyzed by staining with fluorescein-phalloidin (Molecular Probes). The cells were fixed with paraformaldehyde (4% w/v in PBS), permeabilized with phosphatidylcholine, and incubated with FITC-phalloidin according to the instructions of the manufacturer. The cell content of F-actin was determined by flow cytometry.
Association of the ligand-receptor complex with the cytoskeleton
To determine the amount of receptor-ligand complex associated with the cytoskeleton, the cells were allowed to interact with FMLP at 15°C. Interaction with the peptide at 15°C permits binding of the ligand-receptor complexes to the cytoskeleton, but neither the mobilization of intracellular organelles nor the internalization of the receptor-ligand complex occurs (14, 24). Cytoskeletal fractions were prepared as described earlier (24). Briefly, the cells were incubated at 15°C for 5 min, after which [3H]FMLP (2 × 10−8 M) was added, either alone or together with an excess of nonradioactive FMLP (10−5 M), and the incubation was continued for an additional 10 min. The cells were pelleted and resuspended in an ice-cold buffer containing Triton X-100 (1% v/v). The samples were mixed and put on ice for 10 min, after which they were centrifuged for 20 s (9000 × g) in a Beckman microfuge (Beckman Instruments, Fullerton, CA). The Triton X-100-insoluble pellet was washed once in the same medium, and the radioactivity associated with the cytoskeleton was determined (24). No radioactivity was associated with the cytoskeleton (Triton X-100-insoluble pellet) when the peptide was introduced in the system 30 s after solubilization with Triton X-100.
Superoxide anion production and mobilization of neutrophil granules
The chemoattractant peptide FMLP induced a very rapid neutrophil response measured as superoxide anion production (Fig. 1). The maximal production was reached within the first minute, and the response was terminated within 3 to 4 min.
The major part of the neutrophil FMLP-R is stored in subcellular vesicles (secretory vesicles) and mobilizable granules (gelatinase granules and specific granules; for review, see 7 . The mobilization of these intracellular organelles, following an intracellular Ca2+ rise (obtained by ionomycin treatment), subsequently endows the neutrophil plasma membrane with new receptor molecules, as shown for CR3 and the FMLP-R (Fig. 2). Accordingly, ionomycin-treated neutrophils showed a primed response to FMLP (Fig. 1).
Ionomycin treatment induced a rapid (maximal within 1 min) and transient (back to resting value after 5 min) increase in the cellular content of F-actin, revealed by FITC-phalloidin staining (Fig. 3).
Interaction of the receptor-ligand complex with the cytoskeleton determined as desensitization and Triton X-100 insolubility
Neutrophils that were allowed to interact with the chemoattractant FMLP at 15°C were desensitized (Fig. 4), i.e., when these cells were transferred to 37°C, they did not respond to FMLP. This desensitization was stimulus specific. In fact, neutrophils desensitized to FMLP were primed in response to fluoride ions (presumably acting in conjunction with trace levels of aluminum), while the PMA-induced response was unchanged (not shown). We as well as others have shown earlier that binding of FMLP to its receptor at 15°C is associated neither with any granule secretion nor with activation of the NADPH-oxidase (24, 25).
In contrast to what was found when using cells with their vesicles/granules intact (control neutrophils), no desensitization could be induced in ionomycin-treated neutrophils (Fig. 5). This finding suggests that the Ca2+-induced mobilization of vesicle/granule-localized receptors is associated with an impaired binding between the occupied receptor and the cytoskeleton. When the level of cytoskeleton coupling of the occupied receptors was determined by extraction of ionomycin-treated neutrophils with Triton X-100, the number of cytoskeleton-bound receptors (i.e., receptors detected in the detergent insoluble fraction) was indeed reduced, despite the fact that the number of surface-exposed receptors was increased (Fig. 2).
Binding of FMLP at 15°C was not associated with any measurable increase in the cellular content of F-actin in either of the cell populations studied (not shown).
Desensitization in exudated neutrophils
The lack of desensitization (as well as the impaired anchoring of FMLP-occupied receptors to the cytoskeleton) in ionomycin-treated neutrophils could be explained if the receptors mobilized from the storage pools lacked the ability to become desensitized. To test this hypothesis, we determined the ability of exudated neutrophils to become desensitized.
Neutrophils exert their function in vivo mainly after extravasation, a process associated with a hierarchical mobilization of the intracellular storage organelles. In exudated neutrophils, all secretory vesicles are mobilized together with ∼40% of the gelatinase granules and 20% of the specific granules (8). The extravasation process thus results in an increased surface expression of various receptors, including the FMLP-R (19), and as a consequence, such cells were primed with respect to the NADPH-oxidase activity induced by FMLP (Fig. 6).
In disagreement with the results obtained with ionomycin-treated cells, the exudated neutrophils could be desensitized by incubation with FMLP at 15°C (Fig. 7). The hypothesis stated above was thus proved false.
Termination of the FMLP-induced NADPH-oxidase activity
Many FMLP-induced responses (including activation of the NADPH-oxidase) show transient kinetics. It has been suggested that the rapid termination of the responses occurs through the desensitization brought on by the coupling of the receptor-ligand complex to the cytoskeleton (14, 15, 16). To investigate whether this assumption holds true, we studied the kinetics (and thereby the termination) of the response in the primed cell populations (ionomycin-treated and exudated cells, respectively) that differ in ability to become desensitized.
We have shown earlier that the technique used to measure oxidase activity (isoluminol/HRP CL) is well suited for real time studies of superoxide anion release from activated neutrophils (21). In both ionomycin-treated and exudated cells, the FMLP-induced response was increased with respect to magnitude as well as to the duration of the response (Figs. 1 and 6). However, the kinetics of the production of superoxide anion was roughly the same in these two cell populations, irrespective of whether the cells were desensitizable (i.e., exudated cells) or not (i.e., ionomycin-treated cells) (Table I). Hence, our results show that the processes of desensitization and termination of oxidase activity in response to FMLP are not linked.
|.||Cell Populations .||.||.|
|.||Control .||Exudated .||Ionomycin-treated .|
|Tmaxa||0.87 ± 0.15b||1.05 ± 0.07||1.10 ± 0.10|
|ΔTc||1.27 ± 0.12||2.10 ± 0.14||1.93 ± 0.15|
|Integration/peakd||2.53 ± 0.21||2.80 ± 0.14||3.37 ± 0.29|
|.||Cell Populations .||.||.|
|.||Control .||Exudated .||Ionomycin-treated .|
|Tmaxa||0.87 ± 0.15b||1.05 ± 0.07||1.10 ± 0.10|
|ΔTc||1.27 ± 0.12||2.10 ± 0.14||1.93 ± 0.15|
|Integration/peakd||2.53 ± 0.21||2.80 ± 0.14||3.37 ± 0.29|
aTmax represents the peak time (minutes) of the CL response.
bEach numeral represents the mean ± SD, n = 3.
cΔT represents the time difference between the time point of the half-fall of the CL response and the Tmax.
dIntegration/peak represents the integrated value of the CL response divided by the peak CL value.
The mechanism for activation of the neutrophil NADPH-oxidase via receptors for the N-formylated peptide FMLP has been shown to involve a pertussis toxin-sensitive heterotrimeric G protein that can pass the signal on to several transduction pathways working in parallel (10). In general, such signals are highly regulated, allowing cellular adaptation to a wide range of conditions. Desensitization is one such process (occurring in many cell types, including the neutrophils), which is characterized by an absence of response despite the presence of agonist-occupied receptors (26). A model has been put forward stating that the desensitization process may also be of importance for the termination of FMLP-induced neutrophil NADPH-oxidase activity (14, 15, 16), and according to this model, the production of second messengers ceases when the receptor-ligand complex binds to the cytoskeleton. We found that the desensitization phenomenon was absent in ionomycin-primed cells, and in accordance with the described model, the amount of receptor-ligand complexes found to be associated with the cytoskeleton was decreased.
We know from earlier studies (22, 27, 28) that ionomycin induces a mobilization of neutrophil granules to the cell surface, resulting in an increased exposure of different receptors. To measure the amount of FMLP receptors on ionomycin-primed cells, FITC-labeled peptide was allowed to bind to the FMLP receptors, and the cells were analyzed by FACS without prior washing. This technique is superior to the technique previously used (both by us and others), which was based on binding of radioactively labeled peptides, since the latter required sedimentation and washing of the cells to remove unbound peptide. This latter treatment caused a partial detachment of bound ligand, resulting in an underestimation of binding. Furthermore, the characteristics of the binding in that system were such that it did not allow an extrapolation from the binding data of receptor number and binding affinity (29, 30, 31). The receptor binding data that we obtained, showing a very low binding to ionomycin-treated cells, using a protocol with a radiolabeled ligand and a brief washing procedure (not shown), and at the same time a high level of binding with the FITC-labeled ligand, suggest that the number of receptors are increased in these cells but that the dissociation constant is higher compared with the nontreated cells. Using high resolution binding data, it has been shown that the FMLP-R may exist in two interconverting forms (32, 33), one low affinity and one high affinity state. A detailed characterization of the receptor conversion is beyond the specific aims of the present work, which is focused on desensitization and termination of the NADPH-oxidase; however, our results fit with the suggestion that the conversion of the low affinity receptors to the high affinity state is missing in ionomycin-treated cells.
The degree of receptor exposure was higher in ionomycin-primed cells (shown also for the CR3 molecules that are stored in the same compartments as the FMLP-R (7)) compared with untreated control cells. We therefore conclude that the decrease in number of ligand-receptor complexes recovered in the Triton X-100-insoluble phase from ionomycin-primed cells is not due to a decrease in receptor exposure but rather to an impaired association of the ligand-receptor complexes with the cytoskeleton.
One way to fit our data into the model of desensitization would be to suggest that the newly mobilized receptors lack the ability to become desensitized. To investigate this possibility, we determined the FMLP response in neutrophils that had exudated from the blood stream, such cells having mobilized their intracellular pools of receptors (8). The fact that the exudated cells could be desensitized suggests that impaired desensitization is not due to a recruitment of new, nondesensitizable receptors. In addition, our results clearly show that the signals mediating desensitization are also generated in exudated neutrophils, i.e., in cells that have experienced an intimate contact with the vascular endothelium and the basement membrane and have been exposed to a large variety of inflammatory mediators present in the exudate.
The molecular mechanisms behind the lack of desensitization in ionomycin-primed cells can only be speculated upon at this time. It has been shown that an occupation of the receptor for complement fragment 5a triggers an association of the FMLP-R to the cytoskeleton (17), indicating that the association of these receptors to the cytoskeleton is not a process driven by ligand-induced changes of the receptor. It has been suggested that the desensitization is due to an actin-dependent/-driven segregation of the active ligand-receptor complex from the amplifying G protein(s). A rise in intracellular Ca2+ affects (directly or indirectly) the degree of actin polymerization (Refs. 34 and 35; Fig. 3) as well as the subcellular distribution of other cytoskeleton proteins (36, 37), and an attractive explanation for the phenomenon observed in ionomycin-primed cells is that these changes secondarily interfere with/affect the capability of the cytoskeleton to bind the occupied receptors. At the present time, however, we can not exclude the possibility that the receptors as well as the amplifying G protein(s) are affected by the rise in intracellular calcium.
Binding of FMLP to its neutrophil surface receptors results in superoxide anion production by the membrane-localized NADPH-oxidase, but the active oxidase molecules are rapidly deactivated (38). A sustained production of reactive oxygen species is thus a result of continued production of second messengers, inducing a replenishment of a small pool of active oxidase (38). According to the hypothesis described above, association between the occupied receptors and the cytoskeleton regulates the generation of activating signals and by that the termination of the oxidase activity (14, 15, 16, 25). We compared the time course of superoxide anion production in control cells and in the two cell populations that were primed with respect to FMLP-induced generation of reactive oxygen species. The time required to reach the peak value was somewhat longer in the primed cells, as was the time for the half-fall of the response (ΔT in Table I). However, the small changes in the kinetics of the response were not related to the desensitization phenomenon, since they occurred not only in ionomycin-treated cells but also in exudated cells. Taken together, these data clearly show that the termination of the oxidase response to FMLP is not linked to the desensitization process.
Our results support the concept that an association of the ligand-receptor complex with the cytoskeleton is linked to a desensitization of the neutrophils with respect to activation of the oxidase. Desensitization may be achieved through a direct interaction of the receptor-ligand complex with cytoskeleton proteins such as actin (16). A rise in intracellular Ca2+ affects the organization of the cytoskeleton (36, 37, 39, 40), and this should be expected to interfere with the desensitization process. With respect to the role of desensitization in the termination of FMLP-induced oxidase activity, the results presented here strongly imply the existence of an alternative, not yet identified mechanism that is responsible for the termination of the oxidase activity.
This work was supported by The Swedish Medical Research Council, The King Gustaf V Memorial Foundation, The Swedish Association Against Rheumatism, The Anna-Greta Crafoord Foundation, The Fredrik and Ingrid Thuring Foundation, and The Anna-Brita och Arne Lundberg Foundation.
Abbreviations used in this paper: FNLPNTL, formyl-Nle-Leu-Phe-Nle-Tyr-Lys; HRP, horseradish peroxidase; CL, chemiluminescence; SOD, superoxide dismutase; KRG, Krebs-Ringer phosphate buffer containing glucose, Ca2+, and Mg2+; Mcpm, 106 counts per minute.