GEF-H1 Transduces FcεRI Signaling in Mast Cells to Activate RhoA and Focal Adhesion Formation during Exocytosis

When antigen-stimulated, mast cells release preformed inflammatory mediators stored in cytoplasmic granules. This occurs via a robust exocytosis mechanism termed degranulation. Our previous studies revealed that RhoA and Rac1 are activated during mast cell antigen stimulation and are required for mediator release. Here, we show that the RhoGEF, GEF-H1, acts as a signal transducer of antigen stimulation to activate RhoA and promote mast cell spreading via focal adhesion (FA) formation. Cell spreading, granule movement, and exocytosis were all reduced in antigen-stimulated mast cells when GEF-H1 was depleted by RNA interference. GEF-H1-depleted cells also showed a significant reduction in RhoA activation, resulting in reduced stress fiber formation without altering lamellipodia formation. Ectopic expression of a constitutively active RhoA mutant restored normal morphology in GEF-H1-depleted cells. FA formation during antigen stimulation required GEF-H1, suggesting it is a downstream target of the GEF-H1-RhoA signaling axis. GEF-H1 was activated by phosphorylation in conjunction with antigen stimulation. Syk kinase is linked to the FcεRI signaling pathway and the Syk inhibitor, GS-9973, blocked GEF-H1 activation and also suppressed cell spreading, granule movement, and exocytosis. We concluded that during FcεRI receptor stimulation, GEF-H1 transmits signals to RhoA activation and FA formation to facilitate the exocytosis mechanism.


Introduction
Mast cells are tissue-resident immune cells that play an important role in many cellular processes, including wound healing, inflammation, and immune responses [1]. However, they also contribute to allergic disease via hyper-responsive reactions [2]. Mast cells contain numerous cytoplasmic granules that package pre-formed pro-inflammatory mediators and dysregulation of their release during allergic reactions propagate disease. The most potent activation of mast cells by allergens is mediated through the FcεRI signaling pathway, which leads to the robust release of pro-inflammatory mediators stored in granules. This process of regulated exocytosis is called degranulation [3].
Mast cell degranulation is tightly regulated since this process releases potent proinflammatory mediators [3]. Allergen binding to IgE leads to the aggregation of the IgE receptor, FcεRI, on the surface of mast cells. This triggers a downstream signaling cascade via the Lyn-Syk-LAT-PLCγ and the Fyn-Gab2-PI3K signaling pathways [4][5][6][7]. Studies from our lab and others have revealed that Rho GTPases are downstream targets of FcεRI signaling and part of the regulatory mechanism of mast cell degranulation [8][9][10][11]. Rho proteins are monomeric G proteins belonging to the Ras superfamily of GTPases that play diverse roles in many cellular processes, particularly those involved in cytoskeletal dynamics [12]. We have shown that antigen activation of mast cells triggers profound morphological transitions that generate cell protrusions which require Rho GTPase function [10,11]. Generation

RNA Isolation and qPCR
Total RNA was extracted with Trizol (Invitrogen, Waltham, MA, USA), following the manufacturer's instruction. Complementary DNA (cDNA) was synthesized from mRNA using oligo dT primers. Briefly, total RNA was extracted from 3 million cells by adding 1 mL of Trizol, then 200 µL of chloroform followed by centrifugation at 12,000× g for 10 min at 4 • C. RNA was precipitated from the top aqueous phase by adding an equal volume of isopropanol, washed in 70% ethanol, and 5 µg was used to synthesize cDNA using 100 units SuperScript™ II Reverse Transcriptase (Invitrogen, Waltham, MA, USA) and 0.5 µg Oligo (dT)12-18 primer (Invitrogen, Waltham, MA, USA) in a 20 µL reaction. To verify the knockdown effects of GEF-H1 shRNA, qPCR was performed using the SensiFAST™ Probe No-ROX kit (Meridian Bioscience, Cincinnati, OH, USA). The qPCR primers for GAPDH were 5 -ACTCCCATTCTTCCACCTTTG and 5 -CCCTGTTGCTGTAGCCATATT, and for GEF-H1 they were 5 -TGTACCAAGGTCAAGCAGAAG and 5 -GCTCTCTGGTGGTTGTCTTAC. For qPCR, a two-step thermocycling reaction was performed based on the Mastercycler ® ep realplex Real-time PCR System (Eppendorf, Hamburg, Germany). The 2 −∆∆Ct method was used to quantify the mRNA levels with GAPDH as a control [25].

Plasmid Preparation and Transfection of RBL-2H3 Cells
Lifeact-mRuby plasmid was used to label F-actin [26]. P CMV -3xHA-RhoA-G14V cloned in pcDNA3.1+ was obtained from the cDNA Resource Center (cDNA.org). A GEF-H1-RNAi resistant mutant construct (GEF-H1-RNAi-Resi) was cloned for re-introduction experiments after the knockdown of the endogenous GEF-H1 mRNA. Full-length GEF-H1 was cloned from RBL-2H3 cell cDNA using the Phusion polymerase (Invitrogen, Waltham, MA, USA) and the forward and reverse primers, respectively, 5 -TCTAAGCTTGTATGTCTCGGATCG AATCCCT and 5 -AGTGGTACCTTAGCTCTCTGAGGCCGTAG. Full-length GEF-H1 was subcloned into the plasmid pmCherry-C1 after Hind3-Kpn1 digestion. This clone of GEF-H1 was used as a template for GEF-H1-RNAi-Resi cloning. The RNAi-resistant primer was designed as follows: forward: CGGAGAGGCCAGAACCTTTAACGGATCCATTGAGCTC TGTAG, reverse: CTACAGAGCTCAATGGATCCGTTAAAGGTTCTGGCCTCTCCG. These primers contained a BamH1 site (underlined) for subsequent verification. A Phusion PCR was performed according to the site-directed mutagenesis strategy previously described [27]. After the transformation of bacteria, clones were selected that incorporated the BamHI site and subsequently verified by Sanger sequencing. Electroporation was used to transfect RBL-2H3 cells with plasmids [28]. A total of 2 million RBL-2H3 cells were mixed with 10 µg purified plasmid in a 400 µL ice-cold electroporation buffer (137 mM NaCl, 2.7 mM KCl, 1 mM MgCl 2 , 10 mM glucose, 20 mM HEPES, pH 7.4). The cell plasmid suspension was then transferred to a 4 mm electroporation cuvette and pulsed by an electric shock at 250 V voltage and 950 µF capacitance (Harvard Apparatus BTX ECM600 Electro Cell Manipulator). Cells were recovered in a complete medium for 24 h and were used for immunofluorescence or live-cell microscopy.

Microscopy
Immunofluorescence microscopy was used to examine the intracellular distribution of granules, cytoskeletons, or the localization proteins in RBL-2H3 cells. Cells grown on coverglass were fixed with 4% (wt/v) paraformaldehyde (PFA) at room temperature (RT) for 30 min, then permeabilized with 0.2% (v/v) Triton-X100 for 15 min. Cells were blocked with 1% bovine serum albumin (BSA) dissolved in PBS, then incubated with primary antibodies for 2 h at room temperature. Cells were washed 5 times with PBS. Alexa Fluor-conjugated secondary antibodies diluted 1:1000 were used as indicated. Oregon green 488 or Alexa 546 conjugated phalloidin diluted 1:2000 was used to stain F-actin and DAPI (4 , 6-diamidino-2-phenylindol) was used to stain nuclei. Cells were mounted on glass slides with ProLong™ Gold Antifade mounting media (ThermoFisher, Waltham, MA, USA). Images were captured by a Zeiss Observer Z1 microscope (Carl Zeiss, Oberkochen, Germany) with a 63X objective (1.4 NA) and processed using Axiovision 4.8 software.
Live-cell imaging was used to visualize the dynamics of granule trafficking by fluorescence microscopy using Lysotracker Red (ThermoFisher, Waltham, MA, USA) cell morphology transitions in a bright-field [11] and F-actin remodeling F-actin by LifeAct-mRuby [26]. Briefly, previously manipulated RBL-2H3 cells (e.g., Lifeact-mRuby transfected, GEF-H1-depleted or sensitized with anti-DNP-IgE) were grown on round coverslips. Coverslips were placed in an Attofluor chamber (ThermoFisher, Waltham, MA, USA) and growth media was replaced with HTB and placed on a 37 • C-heated microscope stage and objective. Images were captured using a PerkinElmer Ultra-VIEW VoX spinning disk confocal microscope (Waltham, USA) with a 63X objective (1.4 NA) using a 10 s imaging interval. After 1 min of imaging, resting cells were stimulated by the addition of 25 ng/mL of DNP-BSA, and drugs or DMSO were added at the same time. Volocity 6.0 software was used to record and analyze the live-cell videos, which were exported as Window Media files at 10 frames/s.

Analysis of Focal Adhesions
Focal adhesions (FAs) were visualized by immunofluorescence microscopy using vinculin antibodies to label them. We used a method for enrichment of FA staining and quantification that was previously described [33]. Briefly, RBL-2H3 cells were grown on coverslips and then treated with 2.5 mM triethanolamine hypotonic buffer for 3 min at room temperature. Cell bodies were removed by hydrodynamic force using a Waterpik ® Cells 2023, 12, 537 5 of 21 WP-100 Ultra water flosser for 10 s. The Waterpik was set to 3 and the nozzle was held 0.5 cm above cells at a 90 • angle to flush the cells. The cell body and nuclei were removed by washing and the FA fraction remained bound to the coverslips. Next, the FAs were fixed with 4% (wt/v) PFA and then labeled with a 1:100 dilution of vinculin antibody. The coverslips were mounted and fluorescent images of identical exposure were captured. FAs were quantified using ImageJ (National Institutes of Health, USA) by measuring the total fluorescence intensity of stained FAs within an individual cell contour [33].

Cell Size Measurement by ImageJ
RBL-2H3 cells undergo spreading and actin remodeling when stimulated [11,23]; thus, the degree of cell spreading can be regarded as an indicator of mast cell activation [34]. The measurement of cell size was performed using ImageJ to analyze the F-actin outlining the cell periphery. Briefly, the selected area of an RGB image was color-thresholded and cells were then outlined using phalloidin-stained areas. Next, the outlined region was automatically analyzed in ImageJ using the "Analyze" feature with the output values of area, mean, and integrated density.

Statistical Analysis
Quantified data are shown as mean ± s.e.m. (standard error of the mean). Student's t-test was used to identify statistically significant differences between responses from data with two independent variables and one-way ANOVA for parametric data with three or more independent variables. A post hoc analysis by Tukey's HSD (honestly significant difference) test was used to identify pairwise significant differences. Differences in nonparametric data (i.e., cell size) were analyzed by Kruskal-Wallis with a post hoc analysis by Dunn's tests. Statistical analyses were performed using Microsoft Excel Xrealstats Add-in (Real Statistics). p-values < 0.05 were considered statistically significant.

Establishment of a Role for GEF-H1 (ARHGEF2) in Mast Cell Degranulation
Previous studies have shown that Rho proteins, such as Rac1 and RhoA, are involved in mast cell granule exocytosis [9][10][11]35,36]. Rho proteins are activated by Rho guaninenucleotide exchange factors (RhoGEFs), a class of proteins that transduce receptor signaling to downstream Rho protein activation. We hypothesized that GEF-H1 may be a putative RhoGEF involved in regulating mast cell granule exocytosis because it has been shown to activate both Rac1 and RhoA [37][38][39], and it is associated with the exocytosis complex called the exocyst [21,22]. We generated an RBL-2H3 cell line depleted of GEF-H1 (GEF-H1 KD) using lentivirus-mediated shRNA knockdown. qPCR and immunoblot analysis were used to verify depletion of GEF-H1 mRNA and protein, respectively. GEF-H1 mRNA levels were reduced by 81.3% ± 7.3% compared to control cells ( Figure 1A) and immunoblot of lysates showed protein levels were similarly reduced ( Figure 1B). The effect of GEF-H1 KD on mast cell granule exocytosis was examined by degranulation assay. Background levels of exocytosis were similar in all cell lines ( Figure 1C, time 0). However, granule exocytosis was significantly reduced in GEF-H1 depleted cells when antigen-stimulated for 15 min and 30 min ( Figure 1C). These results suggest GEF-H1 may have an important regulatory role in the mast cell granule exocytosis mechanism. . Calculated 2 −ΔΔCt values were normalized relative to the control. A 81.3% reduction in GEF-H1 mRNA was observed in the GEF-H1 KD strain (** p = 0.0090 comparing the reduction in GEF-H1 expression in GEF-H1 KD to the control by one-tailed Student's t-test; n = 3). (B) Immunoblot was used to confirm the reduction in GEF-H1 protein in the GEF-H1 KD strain. (C) Degranulation assays were used to determine the effect of GEF-H1 depletion on mast cell granule exocytosis and were statistically analyzed by one-way ANOVA (p < 0.001; n = three independent blots). Post hoc Tukey's tests revealed GEF-H1 knockdown significantly reduced exocytosis in pairwise comparisons between RBL-2H3 and control strains after 30 min of stimulation (**p < 0.001 and 0.00314 for RBL-2H3 and control strains, respectively).

GEF-H1 Knockdown Results in Reduced Cell Activation and Granule Trafficking
The activation of RBL-2H3 cells by antigen stimulation leads to cell spreading [11,40], which indicates cytoskeletal remodeling is actively occurring. Cell activation also results in the redistribution of secretory granules toward the cell periphery [41,42]. We used immunofluorescence and live-cell microscopy to determine the cellular effects of GEF-H1 depletion on granule trafficking and cell morphology. Control RBL-2H3 cells underwent normal spreading and granules were widely dispersed in the cytoplasm after antigen stimulation (Figure 2A, control). In GEF-H1-depleted cells, antigen stimulation resulted in significantly less cell spreading; however, granules seemed to be well dispersed in the cytoplasm (Figure 2A, GEF-H1 KD). Closer examination revealed that granules were enriched adjacent to the plasma membrane in higher abundance in GEF-H1-depleted cells (Figure 2A, right panels), which is consistent with an observed reduction in degranulation (see Figure 1C). Quantification of cell size showed no difference prior to stimulation; however, GEF-H1-depleted cells showed a significant difference in size after 30 min of antigen stimulation compared to RBL-2H3 and control cells ( Figure 2B). The average area of control cells increased 4.2-fold after stimulation, while GEF-H1-depleted cells increased only 1.9-fold. This supports the notion that GEF-H1 may regulate cell morphology, creating a flattened state with more surface area which facilitates granule exocytosis.
To confirm that the loss of cell spreading and granule distribution can be attributed to the depletion of GEF-H1, we examined whether the reintroduction of GEF-H1 into knockdown cells would rescue these defects. For this, we made an RNAi-resistant construct, GEF-H1-Resi, that was tagged with mCherry. The mCherry-C1 empty vector was used as a control. In control cells, expression of GEF-H1-Resi or mCherry-C1 did not alter cell morphology ( Figure 3A). In GEF-H1-depleted cells, the expression of GEF-H1-Resi restored cell spreading, while the expression of mCherry-C1 did not show any rescue effect ( Figure 3B). Analysis of cell size confirmed that, upon antigen stimulation, depletion of GEF-H1 prevented cell spreading, which was restored to normal levels by expression (A) qPCR was used to quantify levels of GEF-H1 mRNA in control cells infected with the shRNA vector or GEF-H1 shRNA (GEF-H1 KD). Calculated 2 −∆∆Ct values were normalized relative to the control. A 81.3% reduction in GEF-H1 mRNA was observed in the GEF-H1 KD strain (** p = 0.0090 comparing the reduction in GEF-H1 expression in GEF-H1 KD to the control by one-tailed Student's t-test; n = 3). (B) Immunoblot was used to confirm the reduction in GEF-H1 protein in the GEF-H1 KD strain. (C) Degranulation assays were used to determine the effect of GEF-H1 depletion on mast cell granule exocytosis and were statistically analyzed by one-way ANOVA (p < 0.001; n = three independent blots). Post hoc Tukey's tests revealed GEF-H1 knockdown significantly reduced exocytosis in pairwise comparisons between RBL-2H3 and control strains after 30 min of stimulation (** p < 0.001 and 0.00314 for RBL-2H3 and control strains, respectively).

GEF-H1 Knockdown Results in Reduced Cell Activation and Granule Trafficking
The activation of RBL-2H3 cells by antigen stimulation leads to cell spreading [11,40], which indicates cytoskeletal remodeling is actively occurring. Cell activation also results in the redistribution of secretory granules toward the cell periphery [41,42]. We used immunofluorescence and live-cell microscopy to determine the cellular effects of GEF-H1 depletion on granule trafficking and cell morphology. Control RBL-2H3 cells underwent normal spreading and granules were widely dispersed in the cytoplasm after antigen stimulation (Figure 2A, control). In GEF-H1-depleted cells, antigen stimulation resulted in significantly less cell spreading; however, granules seemed to be well dispersed in the cytoplasm (Figure 2A, GEF-H1 KD). Closer examination revealed that granules were enriched adjacent to the plasma membrane in higher abundance in GEF-H1-depleted cells ( Figure 2A, right panels), which is consistent with an observed reduction in degranulation (see Figure 1C). Quantification of cell size showed no difference prior to stimulation; however, GEF-H1-depleted cells showed a significant difference in size after 30 min of antigen stimulation compared to RBL-2H3 and control cells ( Figure 2B). The average area of control cells increased 4.2-fold after stimulation, while GEF-H1-depleted cells increased only 1.9-fold. This supports the notion that GEF-H1 may regulate cell morphology, creating a flattened state with more surface area which facilitates granule exocytosis.  Figure 4A, arrows). Granule tracking analysis revealed that GEF-H1 may affect the velocity of granules, which normally increase after stimulation ( Figure 4B). Depletion of GEF-H1 resulted in a 30% reduction in granule velocity, 0.578 +/− 0.0518 µm/s compared to 0.827 +/− 0.0829 µm/s in control cells. The reduced motility of secretory granules and their peripheral accumulation in GEF-H1-depleted cells is consistent with a reduction in granule exocytosis. To confirm that the loss of cell spreading and granule distribution can be attributed to the depletion of GEF-H1, we examined whether the reintroduction of GEF-H1 into knockdown cells would rescue these defects. For this, we made an RNAi-resistant construct, GEF-H1-Resi, that was tagged with mCherry. The mCherry-C1 empty vector was used as a control. In control cells, expression of GEF-H1-Resi or mCherry-C1 did not alter cell morphology ( Figure 3A). In GEF-H1-depleted cells, the expression of GEF-H1-Resi restored cell spreading, while the expression of mCherry-C1 did not show any rescue effect ( Figure 3B). Analysis of cell size confirmed that, upon antigen stimulation, depletion of GEF-H1 prevented cell spreading, which was restored to normal levels by expression of GEF-H1-Resi ( Figure 3C). This rescue of defects in GEF-H1-depleted cells confirms that GEF-H1 plays a role in regulating cell morphology transitions that occur in stimulated RBL-2H3 cells.  We used live-cell imaging to examine the role of GEF-H1 in the dynamics of cell morphology transitions and granule trafficking during antigen stimulation. Cell morphology was imaged by bright-field microscopy and granules were labeled with LysoTracker Red and imaged by spinning-disk confocal microscopy. Videos show control cells first formed dorsal membrane ruffles then large lamellipodia, causing cells to spread and flattened with granules projecting into the flattened areas (Video S1). GEF-H1 KD cells also formed dorsal ruffles soon after stimulation; however, cells did not form large lamellipodia and did not spread. Granules were found to accumulate at the plasma membrane (Video S2). Still, images extracted from videos show granules accumulating at the periphery of GEF-H1-depleted cells, while few granules accumulated at the plasma membrane in control cells ( Figure 4A, arrows). Granule tracking analysis revealed that GEF-H1 may affect the velocity of granules, which normally increase after stimulation ( Figure 4B). Depletion of GEF-H1 resulted in a 30% reduction in granule velocity, 0.578 +/− 0.0518 µm/s compared to 0.827 +/− 0.0829 µm/s in control cells. The reduced motility of secretory granules and their peripheral accumulation in GEF-H1-depleted cells is consistent with a reduction in granule exocytosis.

RhoA, but Not Rac1, Is a Downstream Target of GEF-H1
GEF-H1 has previously been reported to be a RhoGEF for RhoA and Rac1 [37][38][39]43]. To determine the downstream activation target(s) of GEF-H1 in mast cells, we used a pulldown assay with GST-tagged Rhotekin and PAK1 Rho-binding domain probes that bind to active RhoA-GTP and Rac1-GTP, respectively [31,32]. In control cells, antigen stimulation increased the levels of active Rac1-GTP and RhoA-GTP ( Figure 5A,B, respectively). However, in GEF-H1-depleted cells, antigen stimulation resulted in no increase in active RhoA-GTP levels, while active Rac1-GTP levels increased similar to that observed in control cells ( Figure 5A,B, respectively). These results suggest RhoA activation is the downstream target of GEF-H1, since the knockdown of GEF-H1 prevented the activation of RhoA, but not Rac1, after antigen stimulation.
RhoA regulates the formation of stress fibers in various cells [44,45]; therefore, stress fiber formation can be considered a physiological indicator of RhoA activity. Stress fiber formation was examined in the antigen-stimulated RBL-2H3 control and GEF-H1-depleted cells. There were few stress fibers observed in cells prior to stimulation ( Figure 5C,  panels a and d). However, when antigen-stimulated, control cells formed prominent stress fibers across the cell ( Figure 5C, panels b and c, red arrows), while GEF-H1-depleted cells
To determine the downstream activation target(s) of GEF-H1 in mast cells, we used a pulldown assay with GST-tagged Rhotekin and PAK1 Rho-binding domain probes that bind to active RhoA-GTP and Rac1-GTP, respectively [31,32]. In control cells, antigen stimulation increased the levels of active Rac1-GTP and RhoA-GTP ( Figure 5A,B, respectively). However, in GEF-H1-depleted cells, antigen stimulation resulted in no increase in active RhoA-GTP levels, while active Rac1-GTP levels increased similar to that observed in control cells ( Figure 5A,B, respectively). These results suggest RhoA activation is the downstream target of GEF-H1, since the knockdown of GEF-H1 prevented the activation of RhoA, but not Rac1, after antigen stimulation. tigen stimulation triggered the formation of lamellipodia at the leading edge of control cells (Video S5, Figure S2, top panels) and similarly in GEF-H1-depleted cells (Video S6, Figure S2, bottom panels). This suggests that Rac1 activation is maintained in the absence of GEF-H1. These observations are in agreement with results showing Rac1 activation is maintained (see Figure 5A). Taken together, these results suggest that Rac1 was not a downstream Rho protein regulated by GEF-H1 in RBL-2H3 cells during antigen stimulation.  RhoA regulates the formation of stress fibers in various cells [44,45]; therefore, stress fiber formation can be considered a physiological indicator of RhoA activity. Stress fiber formation was examined in the antigen-stimulated RBL-2H3 control and GEF-H1-depleted cells. There were few stress fibers observed in cells prior to stimulation ( Figure 5C, panels a and d). However, when antigen-stimulated, control cells formed prominent stress fibers across the cell ( Figure 5C, panels b and c, red arrows), while GEF-H1-depleted cells lacked similar stress fiber formations ( Figure 5C, panels e and f). This result supports the conclusion that RhoA activation is controlled by GEF-H1 during antigen stimulation. Stress fiber formation could facilitate the projection of the leading edge of cells for cell spreading, which was significantly reduced by GEF-H1 depletion (see Figure 3). [44,45]. To rule out Rac1 as a possible downstream target of GEF-H1, live-cell imaging was used to visualize the dynamic formation of lamellipodia that occurs during RBL-2H3 stimulation [11]. Live-cell imaging via differential interference contrast (DIC) microscopy showed that membrane ruffling occurred in both the control cells (Video S3, Figure S1, top panels) and GEF-H1depleted cells (Video S4, Figure S1, bottom panels). Furthermore, actin remodeling was directly imaged in live cells using the F-actin probe, Lifeact-mRuby. This showed that antigen stimulation triggered the formation of lamellipodia at the leading edge of control cells (Video S5, Figure S2, top panels) and similarly in GEF-H1-depleted cells (Video S6, Figure S2, bottom panels). This suggests that Rac1 activation is maintained in the absence of GEF-H1. These observations are in agreement with results showing Rac1 activation is maintained (see Figure 5A). Taken together, these results suggest that Rac1 was not a downstream Rho protein regulated by GEF-H1 in RBL-2H3 cells during antigen stimulation.

Expression of Constitutively Active RhoA Bypasses GEF-H1
To further examine whether the effects of GEF-H1 depletion were due to a lack of RhoA activation, we transfected cells with a constitutively active RhoA mutant, RhoA-G14V, to determine if defects could be rescued. Control and GEF-H1-depleted cells were transfected with a 3×HA-tagged RhoA-G14V expressed from a CMV promoter, or empty vector for the control. Cells were either left unstimulated or stimulated for 30 min, and then fixed and stained with anti-HA to mark transfected cells, anti-CD63 to mark granules, and Alexa Fluor 405-phalloidin to show cell morphology. Control cells transfected with either vector or RhoA-G14V resulted in granules that were well dispersed and cells that spread after antigen stimulation ( Figure 6A, upper two rows). In GEF-H1-depleted cells, transfection with RhoA-G14V restored normal granule distribution and cell spreading after antigen stimulation, while transfection with an empty vector did not ( Figure 6A, bottom two rows).
Quantification of cell area showed that control cells were significantly larger than GEF-H1-depleted cells when transfected with an empty vector; however, the ability to spread and increase in size was restored to normal levels by transfection of GEF-H1-depleted cells with RhoA-G14V ( Figure 6B). These results show that defects due to GEF-H1 depletion can be bypassed by expressing constitutively active RhoA and supports the conclusion that RhoA is the downstream target of GEF-H1. Taken together, these data suggest that during antigen stimulation of RBL-2H3 cells, RhoA-GEF-H1 signaling is required for morphological transitions to generate an activated state.

The GEF-H1-RhoA Signaling Axis Regulates Focal Adhesion (FA) Formation
Previous studies have shown that RhoA is a key regulator of focal adhesion (FA) formation [44]. In addition, it was shown that focal adhesion kinase (FAK), a key regulator of FA formation, was activated in antigen-stimulated mast cells [46]. Therefore, we next examined whether FA formation was a downstream target of the GEF-H1-RhoA signaling axis in RBL-2H3 cells.
The effect of the FAK inhibitor, PF-573228, on granule exocytosis was examined by a degranulation assay. PF-573228 had no effect on basal levels of degranulation but showed significant inhibition of degranulation after antigen stimulation ( Figure 7A). PF-573228 also prevented cell spreading and granule dispersion ( Figure 7B). This suggests that inhibition of FA formation by PF-573228 disrupted the cell activation mechanism that leads to granule trafficking during antigen stimulation. The number of FAs formed was examined in unstimulated and stimulated cells by shearing away cell bodies and staining the remaining adherent FAs with vinculin antibody ( Figure 7C). Antigen stimulation resulted in an increase in the intensity of FA staining compared to unstimulated cells, while pretreatment with the FAK inhibitor, PF-573228, reduced FA staining ( Figure 7D). These results are consistent with the requirement of FAs to support granule exocytosis in antigen-stimulated RBL-2H3 cells. FA formation was also analyzed in GEF-H1-depleted cells. Control cells showed a robust increase in FA staining after antigen stimulation, while GEF-H1-depleted cells did not show a comparably robust increase in FAs ( Figure 7E,F). These results show that the depletion of GEF-H1 disrupts the formation of FAs, which suggests that the GEF-H1-RhoA signaling axis may facilitate the generation of an activated mast cell state through FA formation.
RhoA activation, we transfected cells with a constitutively active RhoA mutant, RhoA-G14V, to determine if defects could be rescued. Control and GEF-H1-depleted cells were transfected with a 3×HA-tagged RhoA-G14V expressed from a CMV promoter, or empty vector for the control. Cells were either left unstimulated or stimulated for 30 min, and then fixed and stained with anti-HA to mark transfected cells, anti-CD63 to mark granules, and Alexa Fluor 405-phalloidin to show cell morphology. Control cells transfected with either vector or RhoA-G14V resulted in granules that were well dispersed and cells that spread after antigen stimulation ( Figure 6A, upper two rows). In GEF-H1-depleted cells, transfection with RhoA-G14V restored normal granule distribution and cell spreading after antigen stimulation, while transfection with an empty vector did not ( Figure 6A, bottom two rows).

GEF-H1 Is Activated in Antigen-Stimulated Mast Cells via the FcεRI Signaling Pathway
We hypothesize that GEF-H1 transduces signals from the cell surface receptor, FcεRI, to downstream Rho proteins. Therefore, we next examined whether GEF-H1 activation is linked to the FcεRI signaling pathway. We performed assays for GEF-H1 activation using GST-RhoA-G17A for affinity precipitation [30]. RhoA-G17A is a nucleotide-free mutant of RhoA that has a high binding affinity for RhoA-specific GEFs. RBL-2H3 cell lysates were incubated with GST-Rho-G17A or GST only bound to glutathione resin and GEF-H1 was found to bind selectively to the GST-RhoA-G17A probe and not GST ( Figure 8A). Levels of GEF-H1 binding increased during a time course of antigen stimulation, showing that GEF-H1 activation may be linked to FcεRI signaling ( Figure 8B).
GEF-H1 activation has been shown to occur by two distinct mechanisms: microtubule dynamics and phosphorylation. We and others have shown a prominent role in microtubule dynamics in regulating mast cell granule trafficking and exocytosis [42,[47][48][49]. GEF-H1 was previously reported to be a microtubule-bound RhoGEF [38,50] that may link microtubule remodeling to the activation of Rho proteins [37][38][39]. Active GEF-H1 was shown to be regulated by release from microtubules [38,50]. When we preincubated cells with the microtubule-stabilizing drug, taxol, there was no change in active GEF-H1 levels, while the microtubule-destabilizing drug, nocodazole, induced a small increase in active GEF-H1 levels ( Figure 8C). However, immunoprecipitation of GEF-H1 from RBL-2H3 cells showed no association with tubulin ( Figure 8D). These results suggest that while GEF-H1 activation is linked to FcεRI signaling, it might not rely on microtubule dynamics in mast cells and instead may be regulated by phosphorylation.
Syk is an FcεRI proximal kinase that is essential for mast cell degranulation [51]. The Syk-specific inhibitor, GS-9973, potently inhibited antigen-stimulated degranulation with an IC 50 of~1 nM ( Figure 9A). The Syk inhibitor also showed a dose-dependent inhibition of the morphology transitions associated with RBL-2H3 cell activation ( Figure 9B). The Sykdependent regulation of GEF-H1 activation was demonstrated by the GEF activation assay. Levels of active GEF-H1 increased after antigen stimulation but were significantly reduced when cells were preincubated with 10 µM Syk inhibitor ( Figure 9C). These results indicated that activation of GEF-H1 in antigen-stimulated RBL-2H3 cells was Syk-dependent and thus likely regulated by phosphorylation. While GEF-H1 is known to be activated by phosphorylation [43,50,52,53], whether it is a direct substrate of Syk requires further investigation.

Discussion
Mast cells release potent pro-inflammatory mediators by a highly regulated mechanism of granule exocytosis called degranulation. Aggregation of the IgE surface receptor, FcεRI, results in robust degranulation. Here, we show that the RhoGEF, GEF-H1, is a downstream target of FcεRI signaling involved in regulating processes that facilitate mast cell granule exocytosis. GEF-H1 (also known as ARHGEF2) is a multi-domain protein with a tandem DH-PH (Dbl homology-Pleckstrin homology) domain necessary for Rho protein GTP exchange, an n-terminal C1 domain, which suggests it can be regulated by diacylglycerol, and two coiled domains involved in protein interactions. GEF-H1 has been shown to activate both Rac1 and RhoA [37][38][39]54]. However, we found that RhoA is the primary target of GEF-H1 during antigen stimulation in mast cells ( Figure 5). RhoA activation was found to be deficient in GEF-H1-depleted cells, as was the formation of stress fibers which require RhoA [44]. Rac1 activation was unaffected and downstream functions of Rac, such as the formation of lamellipodia, still occurred in GEF-H1-depleted cells (Videos S2 and S4). Furthermore, the expression of constitutively active RhoA rescued stress fiber formation and exocytosis in GEF-H1-depleted mast cells.
link microtubule remodeling to the activation of Rho proteins [37][38][39]. Active GEF-H1 was shown to be regulated by release from microtubules [38,50]. When we preincubated cells with the microtubule-stabilizing drug, taxol, there was no change in active GEF-H1 levels, while the microtubule-destabilizing drug, nocodazole, induced a small increase in active GEF-H1 levels ( Figure 8C). However, immunoprecipitation of GEF-H1 from RBL-2H3 cells showed no association with tubulin ( Figure 8D). These results suggest that while GEF-H1 activation is linked to FcεRI signaling, it might not rely on microtubule dynamics in mast cells and instead may be regulated by phosphorylation.  Mast cells release potent pro-inflammatory mediators by a highly regulated mechanism of granule exocytosis called degranulation. Aggregation of the IgE surface receptor, FcεRI, results in robust degranulation. Here, we show that the RhoGEF, GEF-H1, is a downstream target of FcεRI signaling involved in regulating processes that facilitate mast cell granule exocytosis. GEF-H1 (also known as ARHGEF2) is a multi-domain protein with a tandem DH-PH (Dbl homology-Pleckstrin homology) domain necessary for Rho protein GTP exchange, an n-terminal C1 domain, which suggests it can be regulated by diacylglycerol, and two coiled domains involved in protein interactions. GEF-H1 has been shown to activate both Rac1 and RhoA [37][38][39]54]. However, we found that RhoA is the primary target of GEF-H1 during antigen stimulation in mast cells ( Figure 5). RhoA activation was found to be deficient in GEF-H1-depleted cells, as was the formation of stress fibers which require RhoA [44]. Rac1 activation was unaffected and downstream functions of Rac, such as the formation unstimulated stimulated Rho GTPases are signaling molecules well-known to regulate actin cytoskeletal remodeling in response to extracellular stimuli [44,45]. RhoGEFs, the upstream activators of Rho GTPases, are thus likely to be pivotal signal transducers of external stimuli. Several RhoGEFs have been shown to function in signaling pathways leading to exocytosis in mast cells [15][16][17] and various other secretory cell types [18,19,[55][56][57]. RhoGEFs couple exocytosis with these morphological transitions since their function has been associated with cytoskeleton remodeling that occurs in conjunction with stimulated secretion [58,59]. GEF-H1, in particular, has been shown to be associated with the plasma membrane exocytosis machinery called the exocyst to control exocytosis [21,22]. While the exocyst is likely a universal component of the exocytosis machinery [60,61], it has not yet been demonstrated to be involved in mast cell degranulation. While our results show a GEF-H1 dependence for mast cell degranulation, we could not detect any interactions with the exocyst complex. Hence, the role of Rho GTPase in exocytosis may be specific to cell morphology transitions needed to facilitate granule docking. Indeed, our previous results showed sequential activation of Rac1 first and subsequently RhoA [11]. It is possible that Rac1-stimulated lamellipodia lead to cell spreading, while RhoA-stimulated stress fibers maintain the activated state. Stress fiber formation could also facilitate cell retraction and initiate a recovery phase back to the resting state.
The role of GEF-H1 in mast cell degranulation may also be due to the stimulation of focal adhesions (FAs). Upon stimulation, RBL-2H3 cells flatten and numerous FAs formed in the spreading area (Video S1, Figure 7). We show that FAs are a crucial part of the mast cell exocytosis mechanism as the FA kinase inhibitor, PF-573228, also inhibited mast cell degranulation. Depletion of GEF-H1 led to a reduction in FA formation that occurs after antigen stimulation, suggesting that FA formation may be one of the functions of the GEF-H1-RhoA signaling axis. This is consistent with the GEF-H1 activation of RhoA, as FA formation is driven by RhoA signaling [44,62]. Indeed, a RhoA-GEF-H1 signaling network has been shown to drive localized exocytosis at FA sites [63,64]. The resolution of our images was not sufficient to conclude that exocytosis occurred at FAs and thus a specific role for GEF-H1 in this event remains to be determined.
The activation of GEF-H1 relied on Syk kinase which is part of the kinase cascade activated by the aggregation of IgE-FcεRI complexes. Mast cell activation and degranulation can be effectively blocked by the Syk inhibitor GS-9973; this inhibitor was also found to block GEF-H1 activation (Figure 9). While these data link GEF-H1 activation to the FcεRI signaling pathway, they do not show that GEF-H1 is a direct substrate of Syk. Previous studies have shown that GEF-H1 can be activated by either tyrosine or serine/threonine phosphorylation [43,50,53]. GEF-H1 contains an autoinhibitory domain (AID) with a central tyrosine (Tyr198) surrounded by negatively charged and lipophilic residues, which was proposed to interact with the DH (Dbl homology) domain to block its catalytic activity [53]. GEF-H1 can be activated by Src phosphorylation at Tyr198, leading to the unblocking of the DH domain [53]. This is similar to the activation manner of another RhoGEF; Vav1 can be phosphorylated at Tyr174 to dissociate the DH domain from the AID [65]. GEF-H1 can also be inactivated by phosphorylation. It was shown that the knock-out of the serine/threonine kinase, Pak2, which is highly abundant in mast cells, leads to increased mast cell degranulation [66]. Pak2 phosphorylates GEF-H1 at Ser-885, which induces 14-3-3 binding and its inactivation [66]. This result supports our conclusions that RhoA-GEF-H1 signaling plays an important role in the mast cell granule exocytosis mechanism.
GEF-H1 was previously found to be microtubule-bound GEF. The binding of GEF-H1 to microtubules restricted its GEF activity in various cells, as reviewed in [67]. Nocodazole treatment, which dissociates microtubules, led to the release and activation of GEF-H1 [68,69]. Therefore, there seem to be two modes of GEF-H1 regulation: protein phosphorylation and microtubule-dependent regulation. In RBL-2H3 cells, we found that the localization of GEF-H1 was not markedly altered by microtubule-targeted drugs (data not shown). Treating with nocodazole did result in a slight increase in activated GEF-H1 ( Figure 8C). However, studies in primary mast cells and Jukat T cells suggest microtubule binding was intrinsic to phospho-regulation, such that phosphorylation affected microtubule interaction [66,70]. Our studies relied on the use of RBL-2H3 cells and this model system may have some limitations for the detection of microtubule regulation. Further studies in other cells or animal systems are needed to validate the impacts of microtubule dynamics and protein phosphorylation in the regulation of GEF-H1.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cells12040537/s1, Figure S1: Depletion of GEF-H1 does not block membrane ruffling and lamellipodia formation, Figure S2: Depletion of GEF-H1 does not affect F-actin remodeling at the cell periphery. The following supporting information is available online and can be accessed at the indicated DOIs. Video S1 (https: