The Hemopoietic Rho/Rac Guanine Nucleotide Exchange Factor Vav1 Regulates N-Formyl-Methionyl-Leucyl-Phenylalanine-Activated Neutrophil Functions

Abs, reagents, and buffers

A rabbit polyclonal Ab against Vav1 (SC-132) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse mAbs against Vav1 (05-219) and phosphotyrosine (4G10; 05-321) were purchased from Upstate Biotechnology (Lake Placid, NY). R-PE-conjugated Ab for the formyl peptide receptor (556016) was from BD PharMingen (San Diego, CA). Chemicals were purchased from Sigma-Aldrich (St. Louis, MO) unless otherwise stated. PBS (pH 7.2), distilled water, glycerol, HBSS, and HEPES (125 mM; pH 7.5) were purchased from Life Technologies (Grand Island, NY). Other buffers used in this study include PBS-BG (PBS with 0.1% bovine albumin (BSA) and 1% glucose; pH 7.25–7.4), PBSG (PBS with 0.9 mM CaCl₂, 0.5 mM MgCl₂, and 7.5 mM glucose), and immunoprecipitation lysis buffer (10 mM Tri-Cl (pH 7.5), 50 mM NaCl, 0.5 mM EDTA, 1% Triton X-100, 0.1% BSA, 50 mM NaF, 2 mM Na₃VO₄, 20 μg/ml chymostatin, 2 mM PMSF, 10 μM leupeptin, and 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride). IgG immune complexes were prepared by mixing equal volumes of 625 μg/ml BSA with 6 mg/ml rabbit polyclonal Ab against BSA (ICN Pharmaceuticals, Aurora, OH) for overnight at 4°C with gentle mixing. BSA-Ab aggregates were sonicated for 45–60 s at 20% output, centrifuged for 5 min at 1000 rpm, and then washed and resuspended in PBS.

Animals

Vav1^−/− mice (22) and Rac2^−/− mice (4) were previously generated by targeted gene disruption and backcrossed into the C57BL/6J strain for >10 generations. C57BL/6J mice were purchased from The Jackson Laboratory (Bar Harbor, ME) and used for wild-type controls. Mice were housed in microisolator cages under specific pathogen-free conditions, fed autoclaved food and acidified water ad libitum, and used in experiments at 8–12 wk of age. All studies using mice were approved by the Indiana University Animal Care and Use Committee.

Isolation of bone marrow and peritoneal exudate neutrophils (PEC)

Neutrophils were purified from the bone marrow (BM) and peritoneal cavity by density gradient centrifugation as described (5) except that PBS was used instead of HBSS. The final preparation was kept on ice in PBS (without Ca²⁺, Mg²⁺) until further processing. To elicit PEC, mice were injected with 1 ml of 3% thioglycolate, and peritoneal cells were harvested after 4 and 18 h as described (5).

NADPH oxidase activity

PMA-elicited superoxide production was measured in a quantitative kinetic assay based on the superoxide dismutase-inhibitable reduction of cytochrome c (5). Superoxide production upon FMLP stimulation was measured by an isoluminol chemiluminescence assay (6).

Flow cytometry

The content of filamentous actin (F-actin) in chemoattractant-stimulated neutrophils was measured by flow cytometry using FITC-phalloidin (4, 6). The relative amount of cell surface FMLP receptor expression in unstimulated BM neutrophils was evaluated using a mAb for the murine formyl peptide receptor (4).

Chemotaxis and neutrophil mobilization

A microchemotaxis device was used to assess chemotaxis in vitro (4, 6). Neutrophil mobilization to peripheral blood in vivo was performed similarly to the method described previously (23). Briefly, mice were injected s.c. with either 200 μl of 20 μM FMLP, 250 ng/ml IL-8 (35 nM), or 200 μl of sterile saline. Peripheral blood counts and differentials were determined from tail vein collections obtained 90 min after injection.

Measurement of Vav1 phosphorylation and Rac activation

Vav1 activation was assessed using a standard method based on phosphotyrosine staining of immunoprecipitates isolated from whole-cell lysates using an anti-Vav1 Ab (Santa Cruz Biotechnology). Briefly, 2 × 10⁷ BM neutrophils in a total volume of 0.5 ml of PBSG incubated at 37°C for 2 min before incubation with 5 μl of DMSO, 10 μM FMLP, 250 ng/ml IL-8 (35 nM), or 100 nM leukotriene B₄ (LTB₄) for the indicated times. In some experiments, cells were pretreated with a final concentration of 10 and 100 μM genistein, 1 and 10 μM 4-amino-5-(4-methylphenyl)-7-(t-butyl) pyrazolo[3,4-d]pyrimidine (PP1), 2 and 20 μM LY294002, or 100 nM wortmannin at 37°C for 20 min before stimulation. For analysis of Vav1 phosphorylation after FcγR stimulation, neutrophils were incubated with IgG immune complexes for 1 h at 4°C with gentle shaking, and then for 5 min at 37°C. Following agonist stimulation, 800 μl of ice-cold PBS was added, and cells were pelleted, lysed with immunoprecipitation lysis buffer, and then clarified. Supernatant was incubated with 1 μg of Vav1 Ab overnight, and precipitated with protein A-conjugated Sepharose beads (Amersham, Arlington Heights, IL). No Vav1 was detected in the cell pellet following lysis of either resting or stimulated cells, and there was no evidence of Vav1 lability during the immunoprecipitation procedure (C. Kim and M. C. Dinauer, unpublished observations). Washed immunoprecipitates were electrophoresed on 10% SDS-PAGE. Immunoblots were probed using an anti-phosphotyrosine Ab or Vav1 Ab.

An affinity precipitation assay for Rac activation was performed as previously described (6). The p21-binding domain of p21-activated kinase 3 was expressed as a GST fusion protein using a vector obtained from R. Cerione (Cornell University, Ithaca, NY).

Statistical analysis

Statistical analysis was performed using Instat (GraphPad, San Diego, CA) or Excel (Microsoft, Redmond, WA) software. Data are expressed as mean ± SD unless otherwise indicated. The two-tailed Student’s t test or Mann-Whitney test was used to determine the difference between groups. A value of p < 0.05 was considered significant.

Vav1 is tyrosine phosphorylated in response to FMLP

Activation of Vav1 requires its tyrosine phosphorylation, which is mediated upon ligation of a variety of membrane tyrosine kinase receptors or receptors with linked tyrosine kinases of the Syk/Zap70, Jak, and Src families, including receptors for growth factors, cytokines, integrins, Igs, and Ag presentation (12). In platelets, Vav1 phosphorylation is also stimulated by the thrombin receptor, a seven transmembrane domain receptor coupled to trimeric G proteins (24). We examined whether Vav1 can be activated in murine neutrophils by FMLP and other chemoattractants, which also signal through G protein-coupled receptors. As shown in Fig. 1⇓, the level of Vav1 phosphorylation was increased by 1.5- to 2-fold compared with vehicle-treated controls in neutrophils stimulated with FMLP. Increased Vav1 phosphorylation in FMLP-stimulated neutrophils was detected as early as 10 s of FMLP stimulation and peaked by 60–120 s (Fig. 1⇓B). This increase was similar to that seen in neutrophils stimulated with IgG immune complexes (not shown). However, stimulation with two other neutrophil chemoattractants, IL-8 or LTB₄, did not induce Vav1 phosphorylation over the same time frame (Fig. 1⇓A and data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Tyrosine phosphorylation of Vav in chemoattractant-stimulated neutrophils. Lysates from vehicle- or chemoattractant-treated wild-type BM neutrophils were immunoprecipitated with a polyclonal Vav1 Ab and analyzed by immunoblotting with a p-Tyr Ab (upper panels) or a monoclonal anti-Vav1 Ab (middle panels). Immunoblots are representative of three to four independent experiments. Densitometry of p-Tyr signal normalized to Vav expression as averaged from all experiments in each group is shown in lower panels, except 1C. A, Neutrophils were treated with 10 μM FMLP, 250 ng/ml IL-8, or 100 nM LTB₄ for 60 s at 37°C (n = 4). B, Neutrophils were incubated with either DMSO for 60 s or 10 μM FMLP for 10–600 s at 37°C (n = 4, except for 600 s time point where n = 1). C, Neutrophils were incubated with 10 and 100 μM genistein, 1 and 10 μM PP1, or 2 and 20 μM LY294002 for 20 min before stimulation with 10 μM FMLP for 60 s at 37°C (n = 4). ∗, p < 0.05; ∗∗, p < 0.02, DMSO control vs FMLP stimulation (paired t test).

The neutrophil formyl peptide receptor is coupled to trimeric GTP-binding proteins and transduces signals via multiple pathways that include PI3Ks and the Src-related tyrosine kinases Fgr, Hck, and Lyn (2, 25, 26, 27, 28). To assess pathways that might regulate Vav1 activation, we examined the effects of pharmacologic inhibitors. The Src family kinase inhibitor PP1, which has been shown to inhibit at least Lyn and Hck activation in FMLP-stimulated human neutrophils (27), had the greatest effect on FMLP-induced Vav1 phosphorylation, which was substantially reduced even at the lowest concentration of PP1 tested (1 μM) (Fig. 1⇑C). Incubation of neutrophils with PP1 also inhibited the basal level of Vav1 phosphorylation detected in our neutrophil preparations (not shown). Vav1 phosphorylation in FMLP-stimulated neutrophils was also attenuated in a dose-dependent manner by the tyrosine kinase inhibitor genistein (Fig. 1⇑C). The PI3K inhibitors LY294002 (Fig. 1⇑C) or wortmannin (not shown) produced only a slight decrease in Vav1 phosphorylation.

Vav1-deficient neutrophils have functional defects in FMLP-induced responses in vitro and in vivo

To examine whether Vav1 regulates functional responses of neutrophils to FMLP, we next examined NADPH oxidase activity in Vav1^−/− neutrophils. As shown in Fig. 2⇓A, Vav1^−/− neutrophils isolated from BM had a ≈3-fold decrease in FMLP-elicited superoxide production compared with wild-type cells. NADPH oxidase activity was similarly reduced in FMLP-stimulated PEC (Fig. 2⇓A), indicating that a role for Vav1 in regulating oxidase activity is independent of the activation state of the cell. Cell surface expression of the FMLP receptor was equivalent in resting wild-type and Vav1^−/− neutrophils (not shown), and thus a reduction in FMLP receptor numbers did not account for the defect in FMLP-induced NADPH oxidase activity. In our hands, neither LTB₄ nor IL-8 elicited superoxide production even in wild-type murine neutrophils. The peak rate of superoxide release occurred at ≈30 s after the onset of FMLP stimulation in both wild-type and Vav1^−/− neutrophils, but its amplitude was markedly reduced in Vav1-deficient cells (Fig. 2⇓B). In contrast, Rac2^−/− neutrophils, which have an 8- to 10-fold decrease in overall superoxide production with FMLP stimulation compared with wild-type cells (5, 6), exhibited both a prolonged lag in achieving maximal enzyme rates as well as an overall decrease in activity (Fig. 2⇓B). These results suggest that Vav1 synergizes with or amplifies pathways acting through Rho/Rac GTPases and is required for full activation of FMLP-induced superoxide production, which is also consistent with the potent priming effect of Nef-activated Vav1 on FMLP-stimulated superoxide production in microglial macrophages (19).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

NADPH oxidase activity in FMLP- or PMA-stimulated neutrophils. Superoxide dismutase-inhibitable superoxide anion production was measured in BM polymorphonuclear neutrophils (BM-PMN) or PEC harvested 4 or 18 h after i.p. thioglycolate. ▪, Wild type; □, Vav1^−/−. A, Isoluminol chemiluminescence was used to measure superoxide production in BM-PMN (n = 6) and PEC (n = 4–5) following stimulation with 10 μM FMLP. Data shows relative luminescence units (RLU) integrated over 100 s in wild-type or Vav1^−/− neutrophils. B, Time course of superoxide production in BM-PMN following stimulation with 10 μM FMLP for wild-type, Vav1^−/−, or Rac2^−/− neutrophils, as indicated. C, Superoxide production was measured by cytochrome c reduction assay in BM-PMN (n = 8) and PEC (n = 4–5) following stimulation with 200 ng/ml PMA. Data for A and C are expressed as mean ± SD. ∗, p < 0.001; ∗∗, p < 0.005, wild type vs Vav1^−/− (unpaired t test).

We also examined NADPH oxidase activity elicited by PMA stimulation of Vav1^−/− neutrophils, a potent pharmacologic activator of superoxide production acting via protein kinase C. Both BM and exudate Vav1^−/− neutrophils exhibited a 25–35% decrease in PMA-induced NADPH oxidase activity compared with wild-type cells (Fig. 2⇑C), suggesting that Vav1 is also required for full responsiveness to this stimulus. As for FMLP-elicited NADPH oxidase activity, Rac2-deficient neutrophils have a more severe defect in response to PMA, with ≈75 and 50% reduction in superoxide activity in bone marrow and exudate cells, respectively (4, 5, 6).

We next investigated the role of Vav1 in chemoattractant-induced actin polymerization and migration in vitro. The basal level of F-actin in Vav1^−/− neutrophils was similar to wild-type cells (Fig. 3⇓). However, the rapid generation of F-actin in response to FMLP was markedly reduced in Vav1^−/− BM neutrophils, whereas it was unaffected in response to IL-8 or LTB₄ (Fig. 3⇓). Vav1 has also been shown to regulate actin polymerization downstream of the TCR (12, 13). The selective requirement for Vav1 in FMLP-stimulated F-actin generation correlated with the finding that FMLP induced tyrosine phosphorylation of Vav1, whereas IL-8 or LTB₄ did not (Fig. 1⇑). Transmigration of Vav1^−/− BM neutrophils in the presence of buffer alone was reduced by ≈2-fold compared with wild-type cells (14 ± 11 vs 27 ± 10; n = 8; p < 0.01; Vav1^−/− vs wild type). Migration elicited by 1 and 10 μM FMLP, but not 0.1 μM FMLP, was reduced by an additional 15–25% in Vav1^−/− neutrophils (Fig. 4⇓A). Hence, absence of Vav1 results in impaired FMLP-induced signals that regulate changes in the neutrophil actin cytoskeleton and cell motility, although Vav1^−/− neutrophils exhibited only a small change in directed migration to FMLP.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

F-actin generation by chemoattractant-stimulated neutrophils. Freshly isolated BM neutrophils were warmed at 37°C in HBSS for 3 min before stimulation with 10 μM FMLP (A), 250 ng/ml IL-8 (B), or 100 nM LTB₄ (C). •, Wild type; □, Vav1^−/−. Treatment with carrier (0.5% DMSO for FMLP, PBS for IL-8, and ethanol for LTB₄) did not stimulate F-actin generation. Wild-type and Vav1^−/− cells were analyzed in pairs. Results are expressed as mean channel fluorescence (MCF), with the baseline wild-type fluorescence arbitrarily assigned a value of 100. Mean ± SD of two to four independent experiments per group. ∗, p < 0.02, wild type vs Vav1^−/− (paired t test).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Neutrophil migration in response to FMLP. A, Migration of BM neutrophils was assayed in a modified Boyden chamber using FMLP (10⁻⁷–10⁻⁵ μM). The numbers of cells migrating to the lower side of the filter in chambers containing buffer alone were also counted in samples assayed in parallel (27 ± 10 for wild-type neutrophils and 14 ± 11 for Vav1^−/−; n = 8; p < 0.01; wild type vs Vav1^−/−), and have been subtracted from the data shown for FMLP-treated groups. ▪, Wild type; □, Vav1^−/−. Data are expressed as mean ± SEM. ∗, p < 0.02, wild type vs Vav1^−/− (paired t test). B, Peripheral blood neutrophil concentrations in wild-type or Vav1^−/− mice 90 min following s.c. injection of 200 μl of sterile saline (□), 20 μM FMLP (▪), or 250 ng/ml IL-8 (hatched bars) (n = 5–9 mice in each group). Data are expressed as mean ± SEM. ∗, p < 0.01 (Mann-Whitney test).

Systemic administration of chemoattractants produces an increase in the peripheral blood neutrophil concentration, commonly referred to as mobilization, which reflects release of mature neutrophils from the BM storage pool (29, 30). Mice with targeted disruption of the high affinity N-formyl peptide receptor lack this response (23). The mechanisms underlying chemoattractant-induced neutrophil mobilization, which does not seem to require any known adhesion molecules, are not fully understood but have been speculated to involve chemoattractant-induced changes in the actin cytoskeleton (reviewed in Ref.30). Similar to Rac2-null mice (4), we found a modest increase in baseline peripheral blood neutrophil counts in Vav1^−/− mice (1890 ± 1180 neutrophils/μl; n = 7) compared with wild-type cohorts (706 ± 670; n = 14; p < 0.01; Vav1^−/− vs wild type). Administration of sterile saline alone did not result in a change in peripheral blood neutrophil counts in either wild-type or Vav1^−/− mice (Fig. 4⇑B). Subcutaneous injection of FMLP induced a ∼3-fold increase in the peripheral blood neutrophil concentration in wild-type mice after 90 min, whereas no significant increase was detected in Vav1^−/− mice (Fig. 4⇑B). In contrast, s.c. injection of IL-8 produced a substantial increase in peripheral blood neutrophil counts in both wild-type and Vav1^−/− mice (Fig. 4⇑B). This is consistent with our findings that IL-8 does not induce phosphorylation of Vav1 and that IL-8-induced F-actin generation is intact in Vav1-null mice. These data demonstrate that Vav1 is involved in FMLP-induced neutrophil mobilization in vivo, and indicate that the impairment in Vav1^−/− mice is selective rather than reflecting a more general defect in this response. Also consistent with a selective defect in FMLP-induced neutrophil trafficking in vivo, the number of neutrophils in peritoneal exudate at 4 and 18 h following i.p. injection of thioglycolate, a nonspecific inflammatory stimulus, was similar in Vav1^−/− and wild-type neutrophils (data not shown).

Regulation of cellular signaling by Vav1 is mediated in large part via activation of Rho family GTPases, and Rac and RhoG appear to be its preferred substrates (13, 15). The defects identified above in Vav1-null neutrophils resemble those of Rac2-null neutrophils, although the functional deficiencies in Rac2-null neutrophils are more severe (4, 5, 6). We previously showed that wild-type murine BM neutrophils contain similar amounts of Rac1 and Rac2, and that stimulation with FMLP induced an increase in both Rac1-GTP and Rac2-GTP as measured by an affinity precipitation assay, with the Rac2-GTP levels being ≈4-fold greater than Rac1-GTP (6). We did not observe a detectable difference in the FMLP-induced increase in either Rac1-GTP or Rac2-GTP in Vav1^−/− neutrophils compared with wild-type neutrophils (not shown). However, FMLP is likely to activate other Rac GEFs such as P-Rex1 (9), and we cannot rule out the possibility that there is a small decrease in Rac activation in the absence of Vav1.