Chemoattractant receptor–stimulated F-actin polymerization in the human neutrophil is signaled by 2 distinct pathways

Neutrophils are cells of the innate immune system, which are recruited to sites of acute inflammation by sensing and crawling along chemoattractant gradients. The crawling of the cell is a result of highly synchronized processes of F-actin polymerization, cytoskeleton contraction, and adhesion to the surrounding tissue. These processes are signaled by G-protein–coupled chemokine receptors1 that after chemokine ligation bind the Gα_i subunit of the trimeric G proteins2 and release the Gβγ subunit. The released Gβγ subunit activates phosphoinositide 3-kinase γ3 (PI3Kγ) and phospholipase C-β (PLC-β).4 PI3Kγ produces phosphatidylinositol 3,4,5-trisphosphate,4 and phosphatidylinositol 3,4-bisphosphate,5 whereas PLC-β produces diacylglycerol and inositol 1,4,5-trisphosphate.4 The release of these lipid products signals a variety of responses in the activated cell, including motility. Apparently, the signaling of motility is strongly dependent on PI3Kγ activation but not on PLC-β.4However, there is a PI3Kγ-independent motility that has been demonstrated studying the migration of neutrophils from PI3Kγ-null mice.4 Cell migration is intimately related to cytoskeleton dynamics,6 suggesting that many of the factors affecting migration may also affect cytoskeleton dynamics. Here we use a micropipet assay to measure the rate of extension of single pseudopods from nonadherent human neutrophils during chemoattractant stimulation. The pseudopods are filled with newly polymerized F-actin (Figure 1B), and we use the rate of pseudopod extension as a measure of the overall rate of F-actin polymerization in the pseudopod region.

Venous blood was drawn from consenting, healthy adult donors into vacutainers anticoagulated with EDTA (ethylenediaminetetraacetic acid). Neutrophils were isolated by a one-step density gradient centrifugation on Polymorphoprep (Nycomed, Oslo, Norway), at 500g for 35 minutes at 23°C. The cells were washed once with Ca²⁺/Mg²⁺-containing Hanks balanced salt solution (HBSS; Sigma Chemical, St Louis, MO) and resuspended in a 25% autologous plasma/HBSS solution. The cells were added to the manipulation chamber and used in the experiments.

A single, passive neutrophil at 37°C was held in a supporting pipet with a 5 μm internal diameter. Another pipet with an internal diameter of 1 μm was filled with f-Met-Leu-Phe (fMLP) and positioned 1 μm or less from the cell surface. The chemoattractant solution was blown over the cell, which initiated the formation and extension of a single pseudopod. Pseudopod extension was observed using an inverted Nikon microscope equipped with a 60× oil immersion objective (numerical aperture 1.4). The microscope images were recorded using a CCD camera (Cohu, San Diego, CA). A real-time counter and the chamber temperature were multiplexed onto the recorded images using a multiplexer (Vista Electronics, La Mesa, CA). The recorded images were analyzed with Metamorph Imaging software (Universal Imaging, Downingtown, PA) and the rate of pseudopod extension was calculated from the measured distances and time lapses. For further details on calculation of the rates of pseudopod extension, see supplementary information included with online article.

The used inhibitors were initially diluted in either ethanol or dimethyl sulfoxide at a maximum final concentration of 0.8% or 0.2%, respectively. The presence of ethanol or dimethyl sulfoxide had no effect on the measured rates of pseudopod extension (data not shown). The cells were incubated with the inhibitors at room temperature (23°C) using the following concentrations (except where otherwise indicated): pertussis toxin (1 μg/mL 60 minutes); wortmannin (WTM; 10 nM to 1000 nM, 15 minutes); dibutyryl cyclic-AMP (dBcAMP; 200 μM, 30 minutes) (Sigma); chelerythrine chloride (10 μM, 30 minutes); diphenyleneiodonium chloride (DPI; 10 μM, 30 minutes); PP2 (20 μM, 20 minutes); piceatannol (50 μM, 30 minutes) (Biomol Research Laboratories, Plymouth Meeting, PA); 1L-6-hydroxymethyl-chiro-inositol 2-(R)-2-O-methyl-3-O-octadecylcarbonate (Akt-inhibitor; 10 μM to 40 μM, 30 minutes) (Calbiochem, San Diego, CA); Clostridium botulinum C3 exoenzyme (C3; 20 μg/mL⁻¹, 3 hours); Y-27632 (10 μM, 30 minutes); U-73122 (5 μM and 25 μM, 30 minutes) (Biomol). Control measurements on cells not incubated with inhibitors were performed before and after the measurements on cells with inhibitors. The rates of pseudopod extension from the 2 groups of control measurements were not statistically different.

Single cells were fixed and stained in the experimental chamber. Each cell was manipulated to produce a single pseudopod and was fixed with 4% paraformaldehyde during pseudopod extension. The fixed cell was transferred into solution containing 50 mg/mL monooleoylphosphatidylcholine (Avanti Polar Lipids, Alabaster, AL) and 100 μM Alexa Fluor 488–conjugated phalloidin (Molecular Probes, Eugene, OR) for 10 minutes at room temperature. Finally, the cell was transferred into HBSS and its fluorescence was observed with a SIT camera (Hamamatsu, Bridgewater, NJ). Fluorescent images were taken in successive 0.50 μm planes and the out-of-focus light was removed using the Metamorph Imaging 3D deconvolution algorithm.

The measured rate of pseudopod extension from the surface of initially passive neutrophils is almost constant7 and depends on chemoattractant concentration.8 The pseudopods extend and retract in the continuous presence of chemoattractant.7 We studied the ability of inhibitors for some of the key signaling molecules of migration to modulate the initial rate of pseudopod extension. We used the specific PI3K inhibitor wortmannin,9 which was shown to reduce neutrophil migration to the level of that observed for PI3Kγ-null neutrophils.4 We stimulated pseudopod extension with 10⁻⁷ M fMLP in human neutrophils incubated with increasing concentrations of wortmannin (Figure 1A). Wortmannin decreased the rate of pseudopod extension by up to 80% compared with control. This result demonstrates that the polymerization of 80% of the F-actin in the pseudopod region is PI3Kγ dependent, while the remaining 20% is not. Pertussis toxin completely abolished fMLP-stimulated pseudopod extension, demonstrating that pseudopod extension is signaled by only the G-protein–coupled receptors.

We incubated neutrophils with the PLC-β inhibitor U73122 to test the possibility for PLC-β dependence on the rate of pseudopod extension. The incubation of the cells in 5 μM U73122 had no affect on the rate of pseudopod extension; however, it completely inhibited the release of intracellular calcium transients monitored with the calcium probe fluo-3 (data not shown). When the cells were incubated in 25 μM U73122, the rate of pseudopod extension was reduced by 20%. However, this reduction could have been a result of nonspecific binding of U73122 at high concentrations. It has been shown that at concentrations above 20 μM, U73122 significantly reduces the intracellular superoxide anion concentration.10 Therefore, there was a possibility that the decreased rate of pseudopod extension observed for the cells in the presence of 25 μM U73122 was a result of decreased superoxide anion concentration. We tested this possibility by incubating cells with the NADPH oxidase inhibitor DPI. The rate of pseudopod extension in the presence of 10 μM DPI decreased by 20%, and the incubation of cells with the combination of 10 μM DPI and 1 μM wortmannin abolished pseudopod formation. These results demonstrate that the PI3Kγ-independent pseudopod extension in the neutrophil is dependent on NADPH oxidase.

The activation of the neutrophil by chemoattractants involves NADPH oxidase activation; however, the involvement of NADPH oxidase activation in F-actin polymerization is unclear. Downstream from the activated chemokine receptors the polymerization of F-actin in the neutrophil is signaled by the GTPases Cdc42 and Rac,11which are also key signaling molecules in motility.12Cdc42 binds to the Wiskott-Aldrich syndrome protein WASp, which recruits phosphatidylinositol 4,5-bisphosphate, profilin, and the Arp2/3 complex, to form new barbed ends for actin polymerization.13 Similarly, Rac binds the WASp family protein WAVE, which recruits profilin and Arp2/3 to form new barbed ends.13 RhoA is another GTPase involved in motility.12 In fibroblasts and other cell types it signals stress fiber formation14; however, in the neutrophil RhoA is known to regulate only integrin detachment.15 The guanine exchange factors that activate Cdc42, Rac, and RhoA are not well characterized; however, it is well documented that GTPase activation is PI3K dependent.16 It is also known that the rearrangement of the cellular cytoskeleton by the Rho GTPases RhoA, Cdc42, and Rac, is dependent on the activation of the Src family tyrosine kinases,17 which are key molecules in the signaling of F-actin polymerization by the integrin receptors.18 Src activity is dependent on chemokine receptor activation similar to Rho GTPase activation19 and is regulated by superoxide anion production.20 These findings suggest that tyrosine kinase activation by chemoattractant receptors may provide an alternative signaling pathway for F-actin polymerization. We inhibited the Src family tyrosine kinases with PP2 and the Syk family tyrosine kinases with piceatannol. The inhibition of Syk with 50 μM piceatannol had no affect on the rate of pseudopod extension (data not shown); however, the inhibition of Src with 20 μM PP2 reduced the rate of pseudopod extension by 20%. The simultaneous incubation of cells with 20 μM PP2 and 1 μM wortmannin abolished pseudopod extension, while the incubation of cells with 20 μM PP2 and 10 μM DPI decreased the rate of pseudopod extension by 30% (Figure1C). These results show that the Src family protein tyrosine kinases are involved in the PI3Kγ-independent signaling of F-actin polymerization.

Since the Src signaling of F-actin polymerization is PI3Kγ independent, while the activation of Cdc42 and Rac is dependent on PI3Kγ,16 we hypothesized that Src activation was mediated by RhoA. To test this hypothesis we incubated cells with the RhoA inhibitor C3 and found that the rate of pseudopod extension decreased by 25% (Figure 1C). Similarly, the inhibition of the RhoA kinase ROCK with 10 μM Y-27632 resulted in the decrease of the rate of pseudopod extension by 20%. The simultaneous incubation of cells with C3 or Y-27632 and 1 μM wortmannin abolished pseudopod extension. These results show that RhoA is involved in the PI3Kγ-independent signaling of F-actin polymerization.

We also studied the atypical PKCζ,21Akt/PKB,22 and cAMP,23 which are known to affect neutrophil migration, for their ability to modulate the rate of pseudopod extension. We measured the rate of pseudopod extension in the presence of 10 μM chelerythrine chloride (Figure 1C) and found that the rate of pseudopod extension decreased by 80%. The simultaneous incubation of cells with 10 μM chelerythrine chloride and 1 μM wortmannin also reduced the rate of pseudopod extension by 80%, whereas the incubation of cells with 10 μM chelerythrine chloride and 20 μM PP2 abolished pseudopod extension. Therefore, PKCζ is part of the PI3Kγ-dependent F-actin polymerization.

Akt/PKB is involved in cell polarization22 and therefore was expected to be involved in the signaling of F-actin polarization. The incubation of cells with increasing concentrations of the Akt-inhibitor24 reduced the rate of pseudopod extension by up to 80% (see supplementary information included with online article). The incubation of cells with both 40 μM Akt-inhibitor and 20 μM PP2 abolished pseudopod extension. Thus, Akt/PKB is part of the PI3Kγ-dependent signaling pathway.

The elevation of the intracellular cAMP has been shown to modulate neutrophil migration.25 The effect of cAMP on migration is most probably manifested through protein kinase A activation and RhoA phosphorylation. Based on the results presented here we expected that the elevation of the cAMP would decrease the rate of pseudopod extension. Indeed, the rate of pseudopod extension for cells incubated with 200 μM dibutyryl cAMP decreased by 20%.

Our results demonstrate the existence of 2 distinct pathways for F-actin polymerization during chemoattractant-stimulated lamella extension in the human neutrophil. One pathway is dependent on PI3Kγ activation and downstream is dependent on PKCζ and Akt/PKB (Figure2). This pathway controls the formation of 70% to 80% of the F-actin in the lamella region. The alternative pathway of F-actin polymerization controls 20% to 30% of the newly formed F-actin in the lamella region. This pathway is dependent on the activation of RhoA, ROCK, Src family tyrosine kinases, and NADPH, and is modulated by cAMP (Figure 2). The exact relation between the latter set of signaling molecules is unknown. However, the reduction of the rate of pseudopod extension by 30% for cells incubated with the combined C3, PP2, DPI, and dBcAMP (Figure 1C, last column) suggests that RhoA, ROCK, Src tyrosine kinase, NADPH oxidase, and cAMP belong to the same signaling pathway.