Research ArticleCell Migration

Dynamic regulation of neutrophil polarity and migration by the heterotrimeric G protein subunits Gαi-GTP and Gβγ

See allHide authors and affiliations

Sci. Signal.  23 Feb 2016:
Vol. 9, Issue 416, pp. ra22
DOI: 10.1126/scisignal.aad8163

Control cAMP to control migration

Activation of the G protein–coupled receptors (GPCRs) that stimulate cellular migration generates active G protein α and βγ subunits, which interact with distinct effector molecules. Using a small molecule that activates βγ subunits without activating α subunits in neutrophils, Surve et al. determined that active βγ subunits alone increased the intracellular concentration of the second messenger cAMP so much that the cells stuck to coated surfaces. Active G protein αi subunits balanced this βγ signal, reducing cAMP sufficiently to enable the cells to move.

Abstract

Activation of the Gi family of heterotrimeric guanine nucleotide–binding proteins (G proteins) releases βγ subunits, which are the major transducers of chemotactic G protein–coupled receptor (GPCR)–dependent cell migration. The small molecule 12155 binds directly to Gβγ and activates Gβγ signaling without activating the Gαi subunit in the Gi heterotrimer. We used 12155 to examine the relative roles of Gαi and Gβγ activation in the migration of neutrophils on surfaces coated with the integrin ligand intercellular adhesion molecule–1 (ICAM-1). We found that 12155 suppressed basal migration by inhibiting the polarization of neutrophils and increasing their adhesion to ICAM-1–coated surfaces. GPCR-independent activation of endogenous Gαi and Gβγ with the mastoparan analog Mas7 resulted in normal migration. Furthermore, 12155-treated cells expressing a constitutively active form of Gαi1 became polarized and migrated. The extent and duration of signaling by the second messenger cyclic adenosine monophosphate (cAMP) were enhanced by 12155. Inhibiting the activity of cAMP-dependent protein kinase (PKA) restored the polarity of 12155-treated cells but did not decrease their adhesion to ICAM-1 and failed to restore migration. Together, these data provide evidence for a direct role of activated Gαi in promoting cell polarization through a cAMP-dependent mechanism and in inhibiting adhesion through a cAMP-independent mechanism.

INTRODUCTION

Cell migration is responsible for multiple processes, including tissue formation, wound healing, and immune responses. Directed cell migration, or chemotaxis, is defined as the movement of a cell toward a chemotactic stimulus, and it involves various environmental cues that activate multiple signaling pathways, which lead to coordination and assembly of multicomponent structures and physical regulation both spatially and temporally (1). These pathways drive cell polarization, which results from the protrusion of the leading edge in the direction of the chemotactic gradient, integrin-mediated adhesion, and retraction of the tail at the back of the cell (2). Cells achieve polarization and directional movement in gradients as shallow as 5% across the length of the cell. Indeed, cells can become polarized and migrate in the absence of a chemotactic gradient, although they do so in random directions. Extensive studies of neutrophils and Dictyostelium discoideum indicate that the receptors for chemoattractants are uniformly distributed on the cell surface, and that polarization occurs because of localized activation of downstream signaling components, which result from self-amplifying positive feedback loops at the leading edge coupled with global inhibitory signals that suppress activation at the trailing edge, key features of the local excitation, global inhibition (LEGI) model for directed cell migration (3, 4).

Chemoattractant receptors are heterotrimeric guanine nucleotide–binding protein (G protein)–coupled receptors (GPCRs) that activate G proteins (consisting of Gαi and Gβγ subunits) of the Gi family (Fig. 1A). Active coupling of GPCRs to G proteins induces a conformational change in the Gα subunit, which leads to its exchange of guanosine diphosphate (GDP) for guanosine triphosphate (GTP) (5, 6). GTP binding induces a conformational change in the Gα subunit, which releases the bound Gβγ subunit. Dissociated Gαi-GTP and Gβγ are the active forms of the proteins and they signal independently of each other. Inactivation occurs through the hydrolysis of GTP to GDP by the Gα subunit and the rebinding of Gβγ to Gα-GDP. Both the chemoattractant receptors and the G proteins are uniformly distributed in the plasma membrane of polarized cells (7).

Fig. 1 Gβγ activation alone reduces neutrophil motility.

(A) Diagram of canonical G protein regulation by GPCRs, including chemoattractant receptors. Pi, inorganic phosphate. (B) Mechanism of action of 12155, which binds directly to Gβγ subunits and results in the release of free Gβγ subunits from Gα-GDP without activating the Gα subunit. (C) Gβγ activation reduces the basal motility of neutrophils. Primary mouse neutrophils were treated with vehicle (DMSO), 10 μM 12155, or 1 μM fMLP, and then were tracked for 25 min by microscopy and analyzed by ImageJ software. Tracks of individual neutrophils for each treatment are shown for a single experiment and are representative of three experiments. (D) Data from three experiments as represented in (C) were analyzed with the Chemotaxis and Migration tool from ibidi to determine the velocity (left) and the distance traveled (right) by the indicated cells. Each point represents an individual cell from three separate experiments that were pooled and analyzed as indicated below. Data from 20 cells under each condition were analyzed for statistical significance by one-way analysis of variance (ANOVA) with Bonferroni posttest. *P < 0.05 and ***P < 0.001. (E) 12155 causes a concentration-dependent decrease in basal migration. Mouse neutrophils were stimulated with the indicated concentrations of 12155 and then were tracked and analyzed as described in (C) and (D). Each point represents the average ± SEM for 20 cells from the data shown in fig. S3 (C and D), which were pooled from three independent experiments.

In response to GPCR activation, Gβγ subunits play a prominent role in immune cell migration through the direct activation of phosphatidylinositol 3-kinase γ (PI3Kγ) (811) and guanine nucleotide exchange factors, such as phosphatidylinositol 3,4,5-trisphosphate (PIP3)–dependent Rac exchanger protein (pREX) and ELMO (engulfment and cell motility)/DOCK (dedicator of cytokinesis), which leads to the subsequent activation of the small GTPases RhoG, Rac, and Cdc42 (1215). The Gβγ-dependent activation of PI3Kγ and the subsequent generation of PIP3 set up a positive feedback loop involving Rac and Cdc42 that ultimately results in the polarized accumulation of PIP3, actin polarization, and formation of the leading edge of the cell (1618). The roles of Gαi signaling in these processes have not been well explored, and it has been suggested that the only role of Gαi is to regulate the release of Gβγ subunits (19, 20). Any determination of specific roles for Gαi in chemotaxis is hampered by the fact that perturbations that inhibit Gαi signaling also inactivate obligatory Gβγ signaling. For example, the modification of Gαi by pertussis toxin blocks any interactions between the Gαi-βγ heterotrimer and GPCRs, thereby inhibiting both Gα and Gβγ signaling. Similarly, knockout of specific G protein α subunits, either in mice or with specific short inhibitory RNAs in cell culture, prevents signaling by both Gα and its associated Gβγ subunits.

We previously identified a small molecule (12155) that acutely activates Gβγ subunit signaling by displacing Gα-GDP from Gβγ without activating Gα, which provides a powerful tool with which to determine the specific functions of the Gβγ and α subunits (Fig. 1B). In experiments with this molecule, we previously demonstrated that activation of Gβγ was sufficient to induce directional neutrophil chemotaxis in a Transwell, suspension-based assay (21), supporting the notion that activated Gαi is not needed for directional chemotaxis in cell suspensions in a steep chemotactic gradient. In vivo, the migration of neutrophils is a more complex process, requiring forward protrusion, adhesion to the endothelial substrata, de-adhesion, and diapedesis. To determine the role of signaling by individual G protein subunits in this complex process, we examined stimulated migration in response to uniform application of 12155 on a two-dimensional surface to mimic endothelial adhesion. In this context, we showed that Gαi-GTP signaling was critical for neutrophil polarization and chemotaxis. We present evidence that when it is present in large quantities, cyclic adenosine monophosphate (cAMP) acts as a global inhibitor of cell polarization and that Gαi-GTP is required to dynamically inhibit cAMP production and to suppress cell adhesion.

RESULTS

Uniform Gβγ stimulation inhibits neutrophil motility on ICAM-1–coated surfaces

To understand how Gβγ signaling regulates cell polarization and migration in the absence of Gα-GTP signaling, we compared the migration of primary mouse neutrophils treated with the chemoattractant N-formyl-Met-Leu-Phe (fMLP) or 12155, a molecule that releases free Gβγ from G protein heterotrimers without activating Gα (Fig. 1, A and B). Mouse neutrophils were imaged on slides coated with the integrin ligand intercellular adhesion molecule–1 (ICAM-1), treated uniformly with vehicle [dimethyl sulfoxide (DMSO)], fMLP, or 12155 (without any gradient), and tracked by differential interference contrast (DIC) microscopy. Basal polarization and migration of primary mouse neutrophils on ICAM-1 is a well-established phenomenon that is important for immune surveillance in vivo. Consistent with this, basal mouse neutrophil migration on ICAM-1–coated slides was unaffected by treatment with vehicle (Fig. 1C). This basal polarization of cells was not observed on glass slides coated with bovine serum albumin (fig. S1), indicating that integrin engagement enhances basal cell polarization in the absence of an applied chemoattractant. Application of fMLP increased the velocity and the distance traveled by these cells (Fig. 1, C and D). In contrast, 12155 completely suppressed basal migration (Fig. 1, C to E, fig. S2, and movie S1). This finding contrasts sharply with what was observed in a Transwell assay in which 12155 stimulated directional chemotaxis (21). Thus, activation of Gβγ in the absence of receptor or Gαi activation actively suppressed cell migration on a surface coated with a cell adhesion molecule.

Uniform Gβγ activation promotes the formation of nonpolarized lamellipodia and increases cell adhesion

Inhibition of cell migration could reflect either decreased polarization of cells, increased adhesion, or simply a decreased ability to respond to chemoattractants. In the absence of chemoattractants, neutrophils plated on ICAM-1 exhibited basal polarity with distinct leading and trailing edges (Fig. 2, A and B, and fig. S3), and stimulation with fMLP increased the percentage of polarized cells (Fig. 2, A and B, and fig. S3). Application of 12155, on the other hand, completely suppressed basal cell polarity, resulting in cells that had no distinct trailing edge and had actin-based protrusions uniformly surrounding the cell, resembling a “fried egg” (Fig. 2, A and B, and fig. S3). The percentage of cells displaying this morphology increased with increasing concentrations of 12155 and we calculated a median effective concentration (EC50) of ~10 μM, which is consistent with the affinity of 12155 for Gβγ in vitro being ~3 μM (Fig. 2C) (21). To measure adhesion, neutrophils were plated on ICAM-1–coated plates, treated with vehicle (DMSO), 12155, or fMLP for 5 min, and then washed, and the numbers of cells that detached from the ICAM-1–coated surface were counted. Treatment with 12155, but not fMLP, decreased the number of cells that detached with washing, which suggested that there was an increase in their adhesion (Fig. 2D). Thus, on an ICAM-1–coated surface, uniform activation of Gβγ with 12155 suppressed cell polarity and increased cell adhesion, both of which likely contributed to the lack of motility of these cells.

Fig. 2 Uniform Gβγ activation promotes the formation of nonpolarized lamellipodia and increases cell adhesion.

(A) Uniform Gβγ activation stimulates the formation of nonpolarized circular lamellipodia. Mouse neutrophils were treated with vehicle (DMSO), 10 μM 12155, or 1 μM fMLP, and then were fixed and imaged by DIC microscopy. Images are representative of multiple cells from four independent experiments. Scale bars, 5 μm. (B) Gβγ activation reduces basal cell polarity. Mouse neutrophils were treated with the indicated reagents as described in (A), fixed, stained for actin, and imaged by epifluorescence microscopy. Cells were scored (between 20 and 50) for each independent experiment in a blinded manner to determine the number of polarized cells. Data are means ± SEM of the percentages of cells that were polarized in three independent experiments. (C) Mouse neutrophils were stimulated with the indicated concentrations of 12155 and were analyzed as described in (B) to determine the percentage of cells that exhibited nonpolarized lamellipodia formation. Data are means ± SEM of three independent experiments. (D) Gβγ activation increases cell adhesion to an ICAM-1–coated substrate. Mouse neutrophils were uniformly stimulated as described in (A), and the numbers of cells that became detached from the surface were counted. Adhesion was calculated as inversely proportional to the number of detached cells. Data are means ± SEM of three independent experiments. All data were analyzed by one-way ANOVA with Bonferroni posttest. *P < 0.05, **P < 0.01, ***P < 0.001.

The effects of 12155 on polarity and adhesion require free Gβγ subunit signaling

To demonstrate the specificity of 12155 for Gβγ, we expressed a well-characterized protein-based Gβγ signaling inhibitor, the C terminus of GPCR kinase 2 (GRK2ct) (22), in an HL-60 human promyelocytic leukemia cell line that can be differentiated into neutrophil-like cells. This cell line has the advantage that exogenous proteins can be expressed by the introduction of mammalian complementary DNA expression constructs by nucleofection. The transfection efficiency of these cells was 20 to 40%, so individual cells expressing GRK2ct were identified by cotransfection with a plasmid encoding yellow fluorescent protein (YFP) before analysis. Untransfected cells and transfected cells expressing YFP alone (35 of 41 cells) responded to 12155 by adopting a circular, flattened morphology as described earlier (Fig. 3 and fig. S4). Cells transfected with plasmid encoding GRK2ct did not respond to 12155 (only 4 of 41 cells responded; Fig. 3 and fig. S4). It is likely that this blockade required large amounts of GRK2ct because cotransfected cells with low amounts of YFP (weakly fluorescent), which were assumed to also have decreased amounts of GRK2ct, showed an increased response to 12155 (15 of 17 weakly fluorescent cells responded to 12155; Fig. 3 and fig. S4). These data support the idea that the ability of 12155 to increase adhesion and decrease cell polarity is directly due to its specific ability to release Gβγ subunits.

Fig. 3 The 12155-dependent effects on cell migration are blocked by GRK2ct.

(A and B) HL-60 cells were transfected by nucleofection with plasmid encoding YFP alone or in the presence of plasmid encoding GRK2ct. Transfected cells were selected on the basis of the abundance of YFP. The cells were further segregated on the basis of whether the cells showed high or low fluorescence, which was interpreted as evidence of increased or decreased amounts of GRK2ct, respectively. (A) Representative individual YFP-expressing cells are shown before and after treatment with 12155. (B) Quantitation of data from the experiments shown in (A). After treatment with 12155, YFP-expressing cells were identified and scored (in a blinded manner) for the appearance of the characteristic flattened fried egg or “normal” morphologies. Cells were imaged at ×60 magnification. Experiments are from four separate sets of transfections, with multiple cells examined in each experiment. Data are means ± SEM of pooled data from three independent experiments and analyzed by one-way ANOVA with Bonferroni posttest. ****P < 0.0001 compared to control cells transfected with YFP alone. The proportions of cells that showed a characteristic flattened symmetrical morphology after treatment with 12155 are as follows:YFP alone, 35 of 41 cells; YFP + GRK2ct (high), 4 of 41 cells; and YFP + GRK2ct (low), 15 of 17 cells.

Receptor-independent activation of both Gαi and Gβγ stimulates polarity and migration

Gβγ subunits are the primary mediators of signals downstream of chemotactic peptide receptors in immune cells. The observation that the activation of Gβγ alone by 12155 suppressed polarity and migration indicates that signals beyond Gβγ activation are necessary to establish polarity and migratory capacity. Chemokine receptors and chemoattractant receptors, such as formyl peptide receptor 1 (FPR1), have the potential to activate Gαi, Gα12/13, GRKs, arrestins, and other receptor-associated signaling molecules. On the other hand, 12155 releases free Gβγ subunits without nucleotide cycling or activating other GPCR-directed pathways (21). We hypothesized that the key difference between the GPCR-dependent and the 12155-dependent activation of Gβγ subunits was the generation of GTP-loaded Gαi. To test this, we used the mastoparan analog Mas7, which specifically and directly activates Gi proteins by binding to the Gi heterotrimer and catalyzing the binding of GTP to the Gαi subunit independently of receptor activation (Fig. 4A) (23). Stimulation of neutrophils with Mas7 led to polarization and migration that was indistinguishable from that observed in response to fMLP (Fig. 4, B and C, and movie S2). Cells treated with either fMLP or Mas7 displayed enhanced polarization, and the number of cells displaying polarity was greater than that under basal conditions, whereas cells treated with 12155 had no distinguishable polarity (Fig. 4, D and E, and fig. S5). These data suggest that Gαi activation is both required and sufficient to complement Gβγ-directed signaling to promote cell polarity and migration.

Fig. 4 Direct activation of Gi heterotrimers is sufficient to stimulate cell migration, induce polarization, and reduce adhesion.

(A) Diagram of the mechanism of action of Mas7, a mastoparan derivative. Mas7 directly interacts with Gi heterotrimers and catalyzes nucleotide exchange on Gαi, which leads to the receptor-independent activation of signaling by both Gαi and Gβγ. (B) Activation of Gi heterotrimers stimulates neutrophil migration. Mouse neutrophils were treated with vehicle (DMSO), 10 μM 12155, 2 μM Mas7, or 1 μM fMLP, and then were tracked and analyzed as described in Fig. 1A. Data are from a single experiment and are representative of four experiments. (C) Activation of Gi heterotrimers increases both the speed of neutrophil migration (top) and the distance traveled (bottom). Mouse neutrophils were treated, tracked, and analyzed as described in Fig. 1A. Data are means ± SEM of 20 cells under each condition from three experiments. (D) Activation of Gi heterotrimers induces cell polarization. Mouse neutrophils were treated with the indicated compounds, fixed, and imaged by DIC microscopy. Images are representative of multiple cells from four individual experiments. Scale bars, 5 μm. (E) Mouse neutrophils were treated with the indicated compounds, fixed, stained for actin, and imaged for fluorescence. Images were analyzed in a blinded manner as described in Fig. 2B to determine the number of polarized cells. Data are means ± SEM of the percentages of polarized cells from three independent experiments. All data were analyzed by one-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001.

Active Gαi prevents the 12155-dependent effects on cell polarization and adhesion

To more directly demonstrate a role for active Gαi in cooperating with Gβγ to enable proper cell polarization and chemotaxis, we transfected differentiated HL-60 cells with plasmids encoding wild-type Gαi1 or a constitutively active form of Gαi, Gαi1Q204L. In each case, the cells were cotransfected with plasmid encoding YFP, and individual fluorescent cells were analyzed (Fig. 5). Neither Gαi1Q204L nor wild-type Gαi1 had a noticeable effect on the behavior of unstimulated cells (Fig. 5, fig. S6, and movies S3 to S5). Furthermore, cells expressing Gαi1Q204L apparently responded normally to fMLP (movie S3). On the other hand, when Gαi1Q204L-expressing cells were treated with 12155, they de-adhered from the ICAM-1–coated surface, polarized, projected pseudopods in specific directions, and migrated slowly (Fig. 5, fig. S6B, and movies S4 to S6). Cells transfected with plasmid encoding wild-type Gαi1 responded to 12155 with a nonpolarized adhesive phenotype that was indistinguishable from that of untransfected cells (Fig. 5, fig. S6A, and movie S7). The marked change in the 12155-dependent phenotype upon introduction of the Gαi1Q204L mutant was variable, but none of the cells displayed the characteristic 12155-induced fried egg morphology. Gαi1Q204L did not fully recapitulate the behavior of cells expressing wild-type Gαi1, in that the cells moved more slowly and had a somewhat elongated phenotype; however, full restoration of the wild-type phenotype was not expected because global Gαi1Q204L overexpression would not be expected to restore the spatiotemporal regulation of Gαi signaling by receptors. Overall, these experiments provide evidence for a direct and active role of Gαi-GTP in conjunction with Gβγ to signal downstream and regulate cell migration in a manner that is independent of the participation of Gαi in the G protein cycle and regulating Gβγ release.

Fig. 5 Treatment of Gαi1Q204L-expressing HL-60 cells with 12155 results in cell polarization and migration.

(A and B) HL-60 cells were cotransfected with plasmid encoding YFP and with plasmid encoding either wild-type (WT) Gαi1 or the Gαi1(Q204L) mutant. Transfected cells were selected on the basis of the abundance of YFP. (A) Representative individual YFP-expressing cells are shown before and after a 30-min treatment with either 12155 or fMLP. (B) Quantitation of data from the experiments shown in (A) for 12155 treatment. After treatment, the YFP-expressing cells were identified and scored (in a blinded manner) for the presence of the characteristic flattened fried egg or normal morphologies. Cells were imaged at ×60 magnification. Data are means ± SEM of pooled data from three independent experiments and were analyzed by a Student’s t test. ****P < 0.0001 compared to control transfected cells expressing YFP alone. The proportions of cells that cells showed a characteristic flattened symmetrical morphology after treatment with 12155 are as follows: YFP + WTGαi1, 22 of 29 cells; YFP + Gαi1(Q204L), 2 of 20 cells. Fluor, fluorescent cells.

i-GTP inhibits cAMP production in neutrophils

The most well understood signaling function of Gαi-GTP is to inhibit adenylyl cyclase (AC), thus decreasing the production of the second messenger cAMP. In contrast to most cell types, activation of Gi-coupled receptors in neutrophils stimulates cAMP production through a noncanonical pathway involving type 9 AC (AC9). In this pathway, Gβγ released from Gi activates mammalian target of rapamycin (mTOR), which in turn activates protein kinase C βII (PKCβII), leading to the phosphorylation and activation of AC9 (2426). We hypothesized that concomitant activation of Gαi might counteract the Gβγ-stimulated accumulation of cAMP. To test this, we measured the ability of fMLP, Mas7, or 12155 to regulate cAMP abundance in either mouse or human neutrophils. Consistent with previous studies, fMLP stimulated an increase in cAMP in primary human neutrophils that peaked at 1 min and decayed over the next 4 min (Fig. 6A). Stimulation with Mas7 also led to a moderate increase in the abundance of cAMP that was comparable to that stimulated by fMLP. In contrast, 12155 led to enhanced stimulation of cAMP accumulation compared to that stimulated by either fMLP or Mas7 (Fig. 6A). The 12155-induced increase in cAMP was sustained compared to that by either Mas7 or fMLP, for which cAMP abundance returned to almost basal amounts over 5 min. The concentration dependence of 12155-dependent cAMP accumulation was validated in primary mouse neutrophils (Fig. 6B). These data suggest that activation of Gαi plays a role in regulating the amount of cAMP produced in neutrophils in response to Gβγ activation.

Fig. 6 Stimulation of Gβγ alone results in enhanced and sustained increases in cAMP abundance compared to activation of Gi protein heterotrimers.

(A) Human neutrophils were stimulated for the indicated times with 10 μM 12155, 2 μM Mas7, or 1 μM fMLP. The amount of cAMP generated under each condition was measured as described in Materials and Methods and is presented as the amount of cAMP (pmol) produced per 75,000 cells. Data are means ± SEM of three independent experiments and were analyzed by one-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001. (B) 12155 results in a concentration-dependent increase in cAMP. Mouse neutrophils were stimulated with the indicated concentrations of 12155 for 30 min and were analyzed as described in (A). Data are presented as the amount of cAMP (pmol) produced per 100,000 cells. Data are means ± SEM of three independent experiments.

Inhibition of PKA restores the polarity of 12155-treated mouse neutrophils

Given the large amounts of cAMP generated in response to 12155, we hypothesized that the increased cAMP abundance inhibited migration. The second messenger cAMP is a dynamic regulator of neutrophil chemotaxis. At low abundance, cAMP is promigratory; however, increased concentrations of cAMP cause decreased migration because of reduced pseudopod protrusion, stronger tail adhesion, and reduced retraction (25, 27, 28). cAMP regulates chemotaxis through the activation of cAMP-dependent protein kinase (PKA), which is required for polarity, but excess activation can inhibit polarity (29). During chemotaxis of D. discoideum, the formation of pseudopods is dependent on cAMP-mediated activation of PKA (30). If the ability of 12155 to inhibit migration was due to excess cAMP production, then the inhibition of cAMP signaling might be able to restore migration in 12155-treated cells. We inhibited cAMP-mediated PKA activation with Rp-cAMPs, an analog of cAMP, or with myristoylated protein kinase inhibitor (14–22) amide (myr-PKI). In contrast to cells treated with 12155 alone, cells pretreated with Rp-cAMPs or myr-PKI and then stimulated with 12155 showed strong polarization, with pseudopod formation at the front of the cell and uropod formation at the tail (Fig. 7, A to D, fig. S7, and movie S8). When neutrophils polarize, actin localizes primarily to the front edge of the cells with some actin being found in the tail. Cells treated with DMSO, Mas7, or fMLP had actin staining at the front edge of the cells and in the tail depicting a polarized morphology (Fig. 7D). Cells in which only Gβγ was activated (with 12155) displayed a uniform distribution of actin with no visible polarity (Fig. 7D). In neutrophils treated with myr-PKI and then stimulated with 12155, actin showed a polarized localization similar to that seen in cells treated with fMLP or Mas7 (Fig. 7D). These data suggest that the enhanced production of cAMP when Gβγ activity is not opposed by Gαi-GTP inhibits the polarization of neutrophils, and that this polarity can be restored by inhibiting cAMP-PKA signaling.

Fig. 7 Inhibition of PKA restores polarity to 12155-treated cells.

(A) Mouse neutrophils were preincubated with myr-PKI or DMSO before being treated with the indicated compounds, fixed, and imaged by DIC microscopy. Images are representative of multiple cells from three independent experiments. Scale bars, 5 μm. (B) Mouse neutrophils were preincubated with the indicated inhibitors, treated with the indicated compounds, fixed, and then imaged by DIC microscopy. The images were analyzed in a blinded manner as described in Fig. 2B to determine the numbers of cells with circular uniform pseudopodia. Data are means ± SEM of three independent experiments and were analyzed by one-way ANOVA. ***P < 0.001 compared to cells treated with 12155 alone. (C) Mouse neutrophils were preincubated with inhibitor, treated with the indicated compounds, fixed, stained for actin, and imaged by epifluorescence microscopy. The images were analyzed in a blinded manner as described in Fig. 2B to determine the numbers of actin-polarized cells. Data are means ± SEM of the percentages of actin-polarized cells from three independent experiments and were analyzed by one-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001. (D) PMNs were preincubated with myr-PKI or DMSO, treated with the indicated compounds, fixed, stained for actin with Acti-stain 555 phalloidin, and imaged by confocal fluorescence microscopy or DIC. Images are representative of multiple cells from three independent experiments. Scale bars, 5 μm.

Inhibition of PKA enhances basal migration but does not restore migration or reduce adhesion in 12155-treated neutrophils

Because inhibition of PKA restored polarity in cells treated with 12155 (Gβγ-stimulated cells), we investigated whether it could also restore migration. The inhibition of PKA led to an increase in the speed of migrating cells and the distance of migration compared to those of basal and 12155-stimulated cells (Fig. 8, A and B, and movie S9); however, inhibition of PKA was unable to restore migration in 12155-treated cells, despite restoring cell polarity (Fig. 8, A and B, and movie S10). These cells were unable to migrate possibly because of a tail retraction defect or the inability of the cells to detach from the substrate (movies S8 and S10). We next investigated whether the inhibition of PKA affected the adhesion of these cells. Cells pretreated with or without the PKA inhibitor and then stimulated with 12155 showed increased adhesion to the substrate compared to that of untreated cells or cells treated with PKA inhibitor alone. Epac (exchange protein activated by cAMP), another target of cAMP, is present in very low amounts in neutrophils and does not mediate adhesion in these cells (31, 32), which suggests that cAMP is not involved in the adhesion induced by 12155. Gαi-GTP rescued the adhesion and migration of 12155-treated cells (Fig. 5), indicating that Gαi-GTP regulates adhesion through a cAMP-independent mechanism.

Fig. 8 Inhibition of PKA enhances basal migration but does not restore migration or reduce adhesion in 12155-treated cells.

(A) Inhibition of PKA enhances basal migration, but not in the presence of 12155. Mouse neutrophils were preincubated with 1 μM myr-PKI (m-PKI) or 300 μM Rp-cAMPs, treated with DMSO or 10 μM 12155, and then tracked for 25 min and analyzed by ImageJ software. Tracks of neutrophils treated with the indicated compounds are depicted. Data are from a single experiment and are representative of three experiments. (B) Pooled data from experiments performed as described in (A). The tracked neutrophils were analyzed with the Chemotaxis and Migration tool from ibidi. Data are means ± SEM of 20 cells for each condition from three independent experiments and were analyzed by one-way ANOVA. *P < 0.05, ***P < 0.001. (C) Inhibition of PKA is unable to reduce Gβγ-stimulated adhesion. Mouse neutrophils were uniformly stimulated with DMSO, 1 μM myr-PKI, 300 μM Rp-cAMPs, 10 μM 12155, or 2 μM Sp-cAMPs, as indicated, and the numbers of cells that detached from the bottom of ICAM-1–coated wells were counted. Adhesion was calculated as inversely proportional to the number of detached cells. Data are means ± SEM of three independent experiments and were analyzed by one-way ANOVA. *P < 0.05 and ***P < 0.001 compared to DMSO-treated control cells. (D) Model depicting the role of Gi proteins in neutrophil migration. When neutrophils are stimulated with the chemoattractant fMLP or with Mas7, both Gαi and Gβγ are activated. Gβγ activation stimulates the production of cAMP. In this model, cAMP inhibits cell polarization by activating PKA, but the concurrent activation of Gαi keeps the concentration of cAMP relatively low, perhaps maintaining a cAMP gradient. In parallel, Gβγ stimulates cell adhesion, which is opposed by Gαi-GTP in a cAMP-independent manner.

DISCUSSION

The role of Gαi-GTP in neutrophil migration

The primary phenotype of neutrophils stimulated uniformly with the Gβγ activator 12155 on ICAM-1–coated surfaces was a strong radial activation of pseudopodia formation without polarization, strong adhesion to the substrate, and basal suppression of migration. These data suggest that Gβγ subunit signaling alone regulates multiple aspects of neutrophil biology, but that additional signals are required to achieve a polarized morphology and cell motility. Here, we showed that a key additional signal is the activation of Gαi. We used multiple approaches to implicate Gαi-GTP signaling in cell migration. The marked phenotypic alteration (increased polarization, migration, and de-adhesion) of 12155-treated cells by overexpression of the constitutively active mutant Gαi1Q204L supports a direct role for Gαi-GTP in driving downstream signaling processes. Under basal or fMLP-stimulated conditions, neither Gαi1Q204L nor wild-type Gαi1 appreciably altered neutrophil function, which is likely because Gαi signaling is only required in the context of a Gβγ stimulus, and this is supplied by endogenous Gαi under receptor-stimulated or Mas7-treated conditions. Only under conditions in which free Gβγ was generated in response to 12155 was a role for Gαi-GTP uncovered. This effect was independent of the role of Gαi in regulating the release of Gβγ subunits because Gαi1Q204L would not be expected to regulate Gβγ because of its constitutive activation, whereas wild-type Gαi and Gαi1Q204L did not appreciably modify fMLP-dependent responses.

One of the downstream effects of Gαi-GTP, which we demonstrated here, is to counterbalance the increases in cAMP generated by the Gβγ-dependent activation of AC (Fig. 8D). Liu et al. (25) showed that receptor-stimulated cAMP is low in abundance at the leading edge and is increased at the trailing edge, although the mechanisms that lead to this polarized distribution have not been described. Gαi-GTP may be involved in locally suppressing cAMP production at the leading edge to a range in which the polarity network can be activated, whereas, in turn, cAMP diffuses to the back of the cell to suppress these signals and activates processes involved in the regulation of uropod dynamics.

At a molecular level, several studies support various aspects of this model. First, the apparently surprising result that the activation of Gi-coupled receptors or Gβγ alone causes increases in cAMP abundance is potentially explained by a noncanonical mechanism involving regulation of AC9, which is abundant in immune cells. In this pathway, Gβγ released from Gi heterotrimers stimulates mTOR activity, which in turn activates the PKCβ-dependent phosphorylation of AC9, leading to increased cAMP production (2426). Our data suggest that under conditions of sole Gβγ activation with 12155, cAMP reaches concentrations that suppress all polarity signaling and likely loses spatial regulation (Fig. 8D). Gαi-GTP inhibits AC9 and thus is poised to reduce the local concentration of cAMP through a membrane-delimited process (33).

PKA activation regulates cell polarity and migration through multiple mechanisms (29), and it is well established that PKA plays a role in regulating cytoskeletal assembly, adhesion, and the directed migration of neutrophils (34, 35). PKA localizes at the leading edge of chemotaxing neutrophils (36, 37), and it inhibits migration and polarity through phosphorylation and inhibition of PIP3-dependent Rac exchanger 1 (P-Rex1), which activates Rac at the leading edge (38, 39). In contrast, cAMP-dependent PKA activation is also required for cell motility because blockade of cAMP production prevents neutrophil migration through inhibition of tail retraction (25, 26). Thus, cAMP-PKA signaling can play opposing stimulatory and inhibitory roles in cell migration, and fine-tuning of the cAMP concentration and its spatial distribution is critical for motility.

Our data suggest that the inhibition of PKA activity in 12155-treated cells restores polarity, but not migration, and that the cells remain strongly adhered to the substrate. A possible concern is that pharmacological inhibitors such as PKI completely inhibit PKA activation, and the proper balance and location of PKA activation that can be achieved through Gαi regulation is likely required for proper migration. An alternate target of cAMP, the Rap exchange factor Epac, could be involved in adhesion that would be insensitive to PKA inhibition, but Epac does not play a role in neutrophil adhesion (31, 32), which suggests that the cAMP-dependent regulation of Epac is not involved in neutrophil migration. Because expression of the constitutively active mutant Gαi1Q204L reduced adhesion, these data suggest that Gαi-GTP regulates adhesion through a cAMP-independent mechanism (Fig. 8D) that remains to be defined.

cAMP as the global inhibitor in the LEGI model

Neutrophils have an inherent ability to polarize even in response to a uniform concentration of a chemoattractant (40). During neutrophil chemotaxis, the asymmetric distribution of lipids, actin, and actin-binding proteins is observed, but the upstream regulatory molecules, including GPCRs and G proteins, are uniformly distributed at the leading edge of migrating cells (3, 4, 41). The uniform distribution of these upstream molecules calls for the localization and regulation of signaling pathways downstream of receptor and G protein activation. The exact mechanism by which these signals are regulated and localized is still undetermined. Several laboratories have developed a LEGI model to explain the ability of a cell to remain polarized under uniform stimulation or very shallow gradients of chemoattractant (17, 18, 42, 43). The model proposes that the chemoattractant stimulates both a self-potentiating positive signal and a slower accumulating globally diffusible inhibitor. The global inhibitor can be overcome locally by the positive feedback signal at the leading edge, but it suppresses the positive feedback loop at the rear of the cell. It is well established that PIP3 generation is a key component of the positive feedback loop to establish cell polarity, but the identity of the global inhibitor has not been clearly defined.

cAMP is an ideal global inhibitor candidate because it is highly diffusible, its cellular localization is controllable, and it is stimulated by Gβγ, which simultaneously stimulates PI3Kγ and other positive signals at the leading edge. The kinetics of increases in cAMP abundance is slow relative to that of PIP3 production, and as shown here, the concentration of cAMP can be modulated by the dynamic interplay between Gβγ-dependent stimulation and Gαi-dependent inhibition immediately downstream of GPCR activation.

In summary, using a distinct set of selective reagents, we have dissected the relative roles of Gα and Gβγ signaling and shown a role for Gαi-GTP subunit signaling in the regulation of neutrophil polarization, adhesion, and migration. This role is played in part through modulation of the concentration of cAMP, which we propose is a key global inhibitor generated downstream of receptor stimulation, and through regulation of cell adhesion by a mechanism that requires further investigation.

MATERIALS AND METHODS

Materials

12155 was originally identified in a screen of the National Cancer Institute Diversity Set with the original identifier number NSC12155 (44). 12155 is also known as surfen [1,3-bis(4-amino-2-methylquinolin-6-yl)urea]. The molecule was prepared as 50 mM stocks in DMSO. Surfen, fMLP, DMSO, and fibronectin were purchased from Sigma-Aldrich. Rp-cAMPs and Mas7 were purchased from Calbiochem. Sp-cAMPs and myr-PKI were purchased from Enzo. Mouse ICAM-1 was purchased from Sino Biological.

Isolation of mouse neutrophils

Neutrophils were obtained from the bone marrow of adult C57BL/6 mice. All procedures were performed on ice with ice-cold buffers. Bone marrow was flushed with phosphate-buffered saline (PBS) (pH 7.4), and red blood cells (RBCs) were lysed with ACK (ammonium-chloride-potassium) lysis buffer. The white blood cells were separated from the lysed RBCs by centrifugation at 325g for 3 min. The cells were then counted, and 1 × 108 cells were used for further isolation. Neutrophils were isolated by magnetic immunodepletion with the Neutrophil Negative Selection Kit (StemCell) according to the manufacturer’s protocol.

Isolation of human neutrophils

Blood was drawn from normal healthy donors according to the protocols followed by the University of Rochester and HIPAA (Health Insurance Portability and Accountability Act). Heparinized blood (20 ml) was added onto a gradient separation kit (1-Step Polymorphs, Accurate Chemical & Scientific Corporation). The various types of cells were separated by centrifugation. The cell layer for polynuclear cells was collected and washed, which was followed by lysis of RBCs with a hypotonic 0.1× PBS solution. The remaining cells collected were mostly neutrophils (>90%) as assessed by flow cytometry.

Nucleofection of HL-60 cells

HL-60 cells differentiated in 1.2% DMSO for 4 days were subjected to nucleofection as follows. Cells (5 × 106) were spun down at 80g for 10 min. After removal of the culture medium, the cells were suspended in 100 μl of Ingenio transfection solution (Mirus Bio LLC). This solution was added to 2 μg of DNA (1 μg of a plasmid expressing enhanced YFP and 1 μg of plasmid expressing the appropriate construct or of empty vector). The mixture was transferred to a 0.2-cm cuvette, and the cells were transfected by nucleofection with an Amaxa Nucleofector II Device (Lonza) on the recommended setting (T-019). After transfection, the mixture was immediately diluted with 500 μl of recovery medium (RPMI 1640 containing 10% fetal bovine serum and 1.2% DMSO) and incubated in an Eppendorf tube at 37°C for 30 min. The cells were then transferred to a dish containing 1.5 ml of recovery medium and incubated overnight at 37°C in an incubator with 5% CO2. The next morning, the transfected cells were plated on coverslips coated with 5 μg of fibronectin from bovine plasma (Sigma-Aldrich). Cells were allowed to attach for 30 min and then were washed with 1 ml of RPMI 1640 medium and subjected to treatment and microscopic analysis in Hanks’ balanced salt solution (HBSS) containing 10 mM Hepes (pH 7.4).

Cell migration assays

Cell migration assays were performed as described previously (45), with modifications. Briefly, Millicell EZ SLIDE 8-well glass (Millipore) or 35-mm glass-bottom dishes were coated with mouse ICAM-1 (1 or 5 μg per well, respectively; Sino Biological) and incubated for 2 hours at 37°C. Naïve mouse neutrophils were isolated and placed on the slide in L-15 medium containing glucose (2 mg/ml; CellGro). Cells were allowed to adhere to the bottom of the slide, and nonadherent cells were washed off. Cells were preincubated with inhibitors for 15 to 20 min, treated with the compounds of interest, and then imaged every 10 s for 25 min. Image acquisition was conducted on a DIC-enabled microscope (Nikon or Olympus) coupled to a Hamamatsu or CoolSNAP HQ (Roper Scientific) camera. The magnification used was ×10, ×20, or ×60. The cells were tracked with the manual tracking functionality in ImageJ software, and the tracked cells were analyzed with the Chemotaxis tool (ibidi). All cells that appeared healthy were tracked, and no thresholding criteria were applied. Velocity for each cell was calculated by dividing the total distance moved by the total migration time.

Adhesion assays

Primary mouse neutrophils were allowed to adhere on ICAM-1–coated plates in L-15 medium containing glucose (2 mg/ml). Cells were stimulated with the appropriate compounds for 5 min at 37°C. For inhibitor studies, the cells were pretreated with a given inhibitor for 15 min before being stimulated. The reaction was stopped by transferring the dish to ice, and the dish was tapped to detach loosely adherent cells. The culture medium was collected in a tube, the cells were washed with ice-cold PBS, the wash was combined with the collected culture medium, and the number of cells present was counted. Adhesion was calculated as the inverse of the number of detached cells relative to control.

Cell polarity measurements

Primary mouse neutrophils were allowed to adhere on an ICAM-1–coated plate in L-15 medium containing glucose (2 mg/ml) and then were stimulated with the appropriate compounds at 37°C. For inhibitor studies, the cells were pretreated with a given inhibitor for 15 min before being stimulated. After 5 min, the cells were washed twice with ice-cold PBS and fixed with 4% paraformaldehyde (PFA) for 15 min at room temperature. The cells were washed with ice-cold PBS and permeabilized with PBS containing 0.1% Triton X-100 for 10 min, which was followed by washing with PBS. Acti-stain 555 phalloidin (150 nM; Cytoskeleton Inc.) was added to the cells and incubated for 20 min, and any excess reagent was washed off with PBS. The cells were visualized, and 10 random images were captured with an epifluorescence microscope (Olympus Inc.) with a 40×, 0.75-NA (numerical aperture) air objective. Random images were selected, and the polarized cells were labeled. Data were pooled, and the number of polarized cells was calculated. For Fig. 7D, the coverslips were mounted on a slide with Fluoromount-G (SouthernBiotech) and imaged with the confocal functionality of a FV1000-AOM multiphoton system (Olympus Inc.) equipped with a 60×, 1.35-NA oil immersion lens (Olympus) using 565-nm excitation.

cAMP assays for mouse neutrophils

Briefly, isolated neutrophils were centrifuged and resuspended at 1 × 106 cells/ml in HBSS with calcium, 10 mM Hepes (pH 7.4), and 0.5 mM 3-isobutyl-1-methylxanthine (IBMX). The cells were incubated at room temperature for 10 to 15 min, and 60,000 cells were used per condition for the assay. The cells were then treated with 0.3 to 30 μM 12155 or with DMSO for 30 min at room temperature. The cells were then centrifuged and resuspended in HBSS with calcium, 10 mM Hepes (pH 7.4), and 0.5 mM IBMX, and 10 μl of cells (60,000 cells) was added to a 96-well plate (½ AreaPlate, PerkinElmer). The cells were then incubated with the LANCE Ultra cAMP detection reagents (PerkinElmer), and measurement of cAMP was performed according to the manufacturer’s protocol.

cAMP assays for human neutrophils

Briefly, isolated human neutrophils were resuspended in HBSS with Ca2+, and 75,000 neutrophils were used per well in a 384-well plate. The cells were stimulated for the appropriate times (0.5, 1, 2, 3, or 5 min), and the assay was performed with the cAMP-Glo Max kit (Promega) according to the manufacturer’s procedure.

Statistical analysis

All statistical analysis was performed by one-way ANOVA using the Bonferroni test functionality. Statistical significance is indicated in the figures as follows: *P < 0.05, **P < 0.01, ***P < 0.001; n.s., not significant.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/9/416/ra22/DC1

Fig. S1. Basal polarization of mouse neutrophils is dependent on ICAM-1.

Fig. S2. Activation of Gβγ subunits alone reduces neutrophil motility on an ICAM-1–coated surface.

Fig. S3. Uniform activation of Gβγ subunits prevents neutrophil polarity.

Fig. S4. GRK2ct blocks the effects of 12155 on cell migration and polarity.

Fig. S5. Activation of Gi heterotrimers stimulates neutrophil polarity.

Fig. S6. The Gαi1Q204L mutant, but not wild-type Gαi1, prevents cell flattening and enables cell polarization and migration.

Fig. S7. Inhibition of PKA restores the polarity of Gβγ-activated cells.

Movie S1. Activation of Gβγ subunits alone reduces neutrophil motility on an ICAM-1–coated surface.

Movie S2. Activation of Gi heterotrimers stimulates neutrophil polarity and migration.

Movie S3. Activated Gαi does not affect the basal or fMLP-stimulated behavior of HL-60 cells.

Movies S4 to S6. Treatment of cells containing activated Gαi with 12155 results in polarized extension of pseudopods, cell movement, and decreased adhesion.

Movie S7. Treatment of cells containing wild-type Gαi with 12155 results in cell flattening and depolarization similar to that of untransfected cells.

Movie S8. Inhibition of PKA restores the polarity of Gβγ-activated cells.

Movie S9. Inhibition of PKA enhances the basal migration of mouse neutrophils in the absence of 12155.

Movie S10. Inhibition of PKA does not restore migration in the presence of 12155.

REFERENCES AND NOTES

Acknowledgments: We thank K. Lim, J. Hak Won, and L. Wagner for technical help, and D. Yule and R. Freeman for the use of microscopes. Funding: This work was supported by grants from the NIH/National Institute of General Medical Sciences (R01GM081772 to A.V.S. and P01AI102851 to M.K.). Author contributions: C.R.S. and A.V.S. conceived of and designed the study and wrote the manuscript with input from all of the authors; C.R.S. performed most of the experiments; J.Y.T. and S.M. performed some supporting experiments; and M.K. assisted with the setup of migration assays and the interpretation of results. Competing interests: The authors declare that they have no competing interests.
View Abstract

Navigate This Article