p21-activated kinase regulates mast cell degranulation via effects on calcium mobilization and cytoskeletal dynamics

Abstract

Mast cells are key participants in allergic diseases via activation of high-affinity IgE receptors (FcepsilonRI) resulting in release of proinflammatory mediators. The biochemical pathways linking IgE activation to calcium influx and cytoskeletal changes required for intracellular granule release are incompletely understood. We demonstrate, genetically, that Pak1 is required for this process. In a passive cutaneous anaphylaxis experiment, W(sh)/W(sh) mast cell-deficient mice locally reconstituted with Pak1(-/-) bone marrow-derived mast cells (BMMCs) experienced strikingly decreased allergen-induced vascular permeability compared with controls. Consistent with the in vivo phenotype, Pak1(-/-) BMMCs exhibited a reduction in FcepsilonRI-induced degranulation. Further, Pak1(-/-) BMMCs demonstrated diminished calcium mobilization and altered depolymerization of cortical filamentous actin (F-actin) in response to FcepsilonRI stimulation. These data implicate Pak1 as an essential molecular target for modulating acute mast cell responses that contribute to allergic diseases.

Figure 1

Targeted disruption of the Pak1 allele. (A) Partial restriction map of the native Pak1 gene (genomic locus), the targeting vector replacing the coding sequence of a portion of the N-terminus, including the p21-binding domain (PBD) and the inhibitory domain (ID) with the Neo-resistance gene in the antisense orientation (targeting vector), and the organization of the targeted Pak1 allele (targeted allele). The 1-kb genomic probe used for screening is indicated along with the expected sizes of the wild-type (WT) and targeted HindIII fragments. (B) Genomic Southern blot analysis (left panel). The nontargeted Pak1 allele (10 kb) is visualized in the WT (+/+) mice and the targeted allele (8 kb) is visualized in the Pak1^−/− (−/−) mice. Both bands can be appreciated in the heterozygous (+/−) mice. Western blot analysis (center panel). WT and Pak1^−/− bone marrow–derived mast cell (BMMC) lysates were subjected to immunoblotting with anti-Pak1. The 68-kDa Pak1 protein is present in WT BMMCs and absent in the Pak1^−/− cells. RT-PCR analysis (right panel). Pak1 cDNA was amplified by PCR from BMMCs to generate a 352–base pair fragment (corresponding to base pairs 306-658) in the WT cells, which is absent in the Pak1^−/− cells. GAPDH mRNA in WT and Pak1^−/− BMMCs is also shown.

Figure 2

Characterization of Pak1^−/− bone marrow–derived mast cells. (A) Bone marrow–derived mast cell (BMMC) receptor expression. BMMCs were maintained in culture for 5 weeks, and expression of c-kit and FcϵRI was measured by incubation with antimouse CD 117 (c-kit) PE-conjugated antibody, and anti-DNP monoclonal antibody IgE clone SPE-7 followed by incubation with FITC-conjugated anti–mouse IgE secondary antibody. Double-positive cells (top right quadrant) are mature mast cells, expressing both c-kit and FcϵRI. Data shown are representative of 6 independent lines from each genotype. (Mean WT = 96.1 + 2.3 SEM % vs Pak1^−/− = 95.3 + 1.7 SEM % double-positive cells, n = 6.) (B) IgE-mediated Pak1 activation in BMMCs (representative of 3 independent experiments). IgE-primed BMMCs were stimulated with antigen (DNP) for the indicated times, lysates were precipitated with anti-Pak1 antibody, and Pak1 activity was assayed. (C) Pak1 activation of pS298-MEK1 in Wt and Pak1^−/− BMMCs. IgE-primed BMMCs were stimulated with antigen (DNP) for the indicated times. Cell lysates (“Western blotting”) were subjected to immunoblotting with anti–phospho-S298 MEK1 (top blot) or anti–total MEK1/2 (bottom blot). (D) Effect of IPA-3 treatment in Wt BMMCs. The length of activation and the addition of inhibitor are indicated.

Figure 3

Pak1 is a critical mediator of mast cell degranulation in vitro. Mast cell degranulation was assessed by measuring the release of β-hexosaminidase. (A) IgE-primed WT and Pak1^−/− BMMCs were stimulated with DNP-HSA for 15 minutes. To determine FcϵRI-independent degranulation, cells were alternatively stimulated with calcimycin (A23187). In all conditions, β-hexosaminidase activity was measured in the supernatant and the extent of degranulation is reported as a percentage of total cellular β-hexosaminidase activity. Data are means plus or minus SEM from triplicate samples in 4 independent experiments. *P < .05, WT versus Pak1^−/−, unpaired, 2-tailed, Student t test. (B) Confocal laser scanning microscopy images of CD63-EGFP in BMMCs. CD63-EGFP fusion protein was introduced into WT and Pak1^−/− progenitors by retroviral transduction as described in “Methods.” Incorporation of the expressed CD63-EGFP into secretory vesicle membranes allows visualization of vesicle location. Fluorescence images of CD63-EGFP (left) and DAPI (right) are shown. Images are representative of 3 independent experiments. (C) Wild-type Pak1 was reintroduced into Pak1^−/− BMMCs by lentiviral transduction. Expression of Pak1 in transduced Wt and Pak1^−/− BMMCs. The recombinant and endogenous Pak1 proteins are indicated. (D) Release of β-hexosaminidase was measured after sensitization and antigen stimulation as in panel A. Data are expressed as a percentage of WT degranulation. Proof of phenotypic rescue by reintroduction of Pak1 is shown for a single transduced mast cell line assayed in triplicate.

Figure 4

Calcium responses are altered in Pak1^−/− BMMCs. IgE-primed WT and Pak1^−/− BMMCs were loaded with Ca²⁺-sensitive dye (fura2) and suspended in Ca²⁺-containing medium (A). To detect release from intracellular stores, cells were suspended in Ca²⁺-free media and treated with EGTA prior to stimulation (B). FcϵRI was activated by addition of DNP-HSA and changes in intracellular calcium concentration (i[Ca²⁺]) were measured by spectrophotofluorimetry. (A) Representative experiment of a WT (black) versus Pak1^−/− (red) BMMC sample demonstrating decreased Ca²⁺-bound fura-2 in the Pak1^−/− cells after antigen stimulation (left panel). Changes in i[Ca] after addition of digitonin followed by EGTA are also shown. Average change in i[Ca²⁺] (right panel) expressed as a percentage of WT change in i[Ca²⁺] from 4 independent experiments, *P < .05, WT versus Pak1^−/−, unpaired, 2-tailed, Student t test. (B) Representative experiment of a WT (black) versus Pak1^−/− (red) BMMC sample demonstrating similar responses upon antigen stimulation after depletion of extracellular calcium (left panel). Average percentage increase in i[Ca²⁺] (right panel) from baseline in WT and Pak1^−/− BMMCs from 3 independent experiments, P < .0.57, WT versus Pak1^−/−, unpaired, 2-tailed, Student t test. (C) Western blot and densitometry for activated PLCγ1. NS indicates cells at baseline (nonprimed, nonstimulated). IgE-primed WT and Pak1^−/− BMMCs were stimulated with DNP for 30 seconds or 1 minute. Cell lysates (“Western blotting”) were subjected to immunoblotting with anti–phospho-PLCγ1 (top blot) or anti–total PLCγ1 (bottom blot) and densitometry was performed. Intensity of phosphorylated PLCγ1 bands is represented as the ratio of phospho/total PLCγ1 for 1 of n = 4 experiments (P = .79 and P = .47 at 30 seconds and 60 seconds, respectively).

Figure 5

Pak1 is required for normal allergen-induced cytoskeletal changes and vesicle trafficking. In all panels, cells were imaged at baseline (NS), after IgE-priming (4 hours), or following IgE-priming and DNP-HSA stimulation (5 minutes) as indicated. (A) WT and Pak1^−/− BMMCs were fixed and stained with Alexa-488 phalloidin (to detect F-actin intensity) and DAPI nuclear stain. The images are representative of 5 experiments. (B) The number of cells showing F-actin disassembly was estimated under indicated conditions using a quantitative intensity analysis of Image J. Data are expressed as a percentage of cells with fragmented rings (= number of cells with fragmented ring/total number of cells imaged × 100) for 4 independent experiments (100 cells counted per experiment); *P < .001, WT versus Pak1^−/−, unpaired, 2-tailed, Student t test. (C) Confocal laser scanning microscopy images of WT or Pak1^−/− BMMCs transduced with pCL1EGFP (control, “−”) or pCL1EGFP-PAK1 (Pak1 transgene, “+”) as indicated. Imaged cells have been treated with IgE and DNP.

Figure 6

Genetic disruption of Pak1 diminishes PCA in vivo. Evans blue extravasation after antigen challenge. (A,B) Pak1^−/− or WT mice (n = 5 genotype) were sensitized by intradermal injection of anti-DNP IgE (1:44 dilution, 1 μg/mL) into the right ear and with PBS into the left ear. After 20 hours, mice were challenged by intravenous injection of antigen (DNP-HSA) in PBS/Evans blue. IgE-primed (right) ears 30 minutes after antigen challenge are shown qualitatively (bottom). From each ear, Evans blue was extracted and the intensity of the dye was measured by absorption at 620 nm. *P < .05, WT versus Pak1^−/−, unpaired, 2-tailed, Student t test. (C) Mast cell–deficient W^sh/W^sh mice (n = 5) were reconstituted locally in the ears by intradermal injection of WT or Pak1^−/− mast cells. The mice were then sensitized by intradermal injection of anti-DNP IgE into the right ear and with PBS into the left ear as in panel A. IgE-primed (right) ears from mice reconstituted with WT or Pak1^−/− mast cells 30 minutes after antigen challenge are shown qualitatively (bottom). From each ear, Evans blue was extracted and the intensity of the dye was measured by absorption at 620 nm. *P < .05, WT versus Pak1^−/−, unpaired, 2-tailed, Student t test.