Assessment of roles for the Rho-specific guanine nucleotide dissociation inhibitor Ly-GDI in platelet function: a spatial systems approach

Abstract

On activation at sites of vascular injury, platelets undergo morphological alterations essential to hemostasis via cytoskeletal reorganizations driven by the Rho GTPases Rac1, Cdc42, and RhoA. Here we investigate roles for Rho-specific guanine nucleotide dissociation inhibitor proteins (RhoGDIs) in platelet function. We find that platelets express two RhoGDI family members, RhoGDI and Ly-GDI. Whereas RhoGDI localizes throughout platelets in a granule-like manner, Ly-GDI shows an asymmetric, polarized localization that largely overlaps with Rac1 and Cdc42 as well as microtubules and protein kinase C (PKC) in platelets adherent to fibrinogen. Antibody interference and platelet spreading experiments suggest a specific role for Ly-GDI in platelet function. Intracellular signaling studies based on interactome and pathways analyses also support a regulatory role for Ly-GDI, which is phosphorylated at PKC substrate motifs in a PKC-dependent manner in response to the platelet collagen receptor glycoprotein (GP) VI–specific agonist collagen-related peptide. Additionally, PKC inhibition diffuses the polarized organization of Ly-GDI in spread platelets relative to its colocalization with Rac1 and Cdc42. Together, our results suggest a role for Ly-GDI in the localized regulation of Rho GTPases in platelets and hypothesize a link between the PKC and Rho GTPase signaling systems in platelet function.

platelets are anucleate cell fragments of megakaryocytes that are well known as the primary cellular mediators of hemostasis (20, 24, 51, 54). Circulating quiescently in a discoid shape, platelets undergo drastic changes in morphology on detection of vessel damage and aggregate with other platelets to protect the integrity of the vasculature (4). After their attachment to extracellular matrix proteins, platelets reorganize their cytoskeleton, leading to the formation of actin-rich filopodia, which are later filled in to make sheets of lamellipodia. The actin regulatory Rho GTPase proteins Cdc42, Rac1, and RhoA all play critical roles in platelet responses that ultimately allow for the formation of a hemostatic plug (10). For instance, Rac1 facilitates the formation of lamellipodia to support platelet spreading as well as thrombus stabilization (2, 39, 41, 43). Cdc42 also has described roles in filopodia formation, platelet spreading, and granule secretion (3, 9, 41, 42). RhoA mediates actin stress fiber formation, actin contractility, and platelet shape change on activation and is essential for the stabilization of thrombi under shear (44). In addition, more recently characterized molecules such as RhoG also have been shown to regulate platelet secretion and thrombus formation (26, 35).

The diverse functional outputs of Rho GTPases are mediated by a cycling between guanosine triphosphate (GTP)–bound “active” and guanosine diphosphate (GDP)–bound “inactive” states, as regulated by Rho GTPase-activating proteins (RhoGAPs), which inactivate Rho GTPases by accelerating the hydrolysis of GTP to GDP, and Rho guanine nucleotide exchange factors (RhoGEFs), which facilitate the exchange of GDP for GTP to drive Rho GTPase activation (10). In addition to GAPs and GEFs, a set of Rho-specific guanine nucleotide dissociation inhibitor proteins (RhoGDIs) also bind to Rho GTPases to block GDP dissociation from Rho GTPases, maintaining the inactive state while anchoring Rho GTPases in specific intracellular locations, serving as an “invisible hand” in the regulation and coordination of Rho GTPase activities intracellularly (25). In mammalian cells, Rho GTPase activities are coordinated by ~70 Rho GAP proteins and 60 GEFs, whereas only three RhoGDI genes are expressed in mammalian cells to mediate Rho GTPase sequestration (30, 31). RhoGDI1 (also known as RhoGDIα or ARHGDIA—referred to here as “RhoGDI” in this study) is ubiquitously expressed, interacts with a number of Rho GTPases, and is considered to be the best characterized RhoGDI family member (17). RhoGDI2 (also known as Ly-GDI, RhoGDIβ, D4-GDI, or ARHGDIB—referred to as “Ly-GDI” in this study) is highly expressed in lymphocytic and hematopoietic cells and is typically found to be more closely associated with the cytoskeleton (22, 25). Loss of RhoGDI activities has multiple functional effects on different cell types, ranging from decreased rates of spreading in mesangial and melanoma cells (16) to increased migration in cancer cells and increased permeability of the pulmonary vasculature (27), consistent with increased RhoA and Rac activities. RhoGDIs are regulated by a number of modalities, including reversible, posttranslational phosphorylation by Src family kinases, protein kinase C (PKC), and p21-activated kinases (25, 28, 31). RhoGDI localization and function is also regulated by prenylation modifications (31). More recently, reversible lysine acetylation has also been demonstrated to regulate differential aspects of RhoGDI function (38).

Despite the critical roles of Rho GTPases and their regulation in platelet function, many relevant Rho GTPase regulatory proteins such as RhoGDIs remain uncharacterized for roles in platelet Rho GTPase regulation and platelet function (10). Here we investigate the expression, localization, and regulation of RhoGDI proteins in human platelets, uncovering the GDI family member Ly-GDI as a putative target of protein kinase C (PKC) in the organized regulation of Rho GTPase proteins in platelets. Our data suggest that Ly-GDI may serve as a cytoskeletal-localized regulator of Rac1 and Cdc42 at specific intracellular locations in a PKC-dependent manner, offering Ly-GDI as a novel regulatory protein relevant to developing models of platelet cytoskeletal signaling in hemostatic function and thrombus formation.

MATERIALS AND METHODS

Reagents.

All reagents were from Sigma-Aldrich except as noted. Prostacyclin (PGI₂) was from Cayman Chemical (Ann Arbor, MI). For static adhesion assays, fluorescence microscopy, and immunoprecipitation, human fibrinogen (FIB3) was from Enzyme Research (South Bend, IN). Anti-Cdc42 (610928), D4-GDI (556498), PKC (610107), and Rac1 (610650) were from BD Biosciences (San Diego, CA). Anti-RhoGDI (sc-360), P-selectin (sc-271267), RhoA (26C4), integrin β₃ (sc-51738), mouse IgGs (sc-2025), rabbit IgGs (sc-2027), and protein A/G PLUS-agarose beads (sc-2003) were from Santa Cruz Biotechnology (Dallas, TX). Phospho-(Ser) pan-PKC substrate antibody (2261) was from Cell Signaling (Danvers, MA). tetramethylrhodamine (TRITC)-phalloidin and the α-tubulin (T6199) antibody were from Sigma. Integrilin (eptifibatide) was from Merck & Co. (Whitehouse Station, NJ). Alexa Fluor secondary antibodies were from Life Technologies (Carlsbad, CA). Collagen-related peptide (CRP) was from R. Farndale (Cambridge University, Cambridge, UK). Ro 31-8220 and Go 6976 were from Tocris (Bristol, UK). Chariot reagent was from Active Motif (Carlsbad, CA).

Preparation of human washed platelets.

Blood was drawn by venipuncture from healthy human donors in accordance with an Oregon Health & Science University IRB-approved protocol into a final concentration of 0.38% sodium citrate as previously described (8, 13). Warmed acid-citrate-dextrose (85 mM sodium citrate, 100 mM glucose, 71 mM citric acid, 30°C) was added to the anticoagulated whole blood and then centrifuged at 200 g for 20 min to separate platelet-rich plasma. The platelet-rich plasma was further purified by centrifugation in the presence of PGI₂ (0.1 μg/ml) at 1,000 g for 10 min. Purified platelets were resuspended in modified HEPES/Tyrode buffer (129 mM NaCl, 0.34 mM Na₂HPO₄, 2.9 mM KCl, 12 mM NaHCO₃, 20 mM HEPES, 1 mM MgCl₂, pH 7.3; supplemented with 5 mM glucose) containing 0.1 μg/ml PGI₂. Platelets were washed once by centrifugation at 1,000 g for 10 min and resuspended in HEPES/Tyrode buffer at indicated concentrations.

Static adhesion assays.

For platelet spreading experiments, 12-mm no. 1.5 glass coverslips (Fisher Scientific or Warner Instruments) or cover glass-bottom dishes (MatTek) were coated with human fibrinogen (50 μg/ml) followed by surface blocking with filtered fatty acid-free BSA (5 mg/ml). Vehicle (0.1% DMSO) or inhibitors (Ro 31-8220, 10 μM; Go 6976 1 μM) were added to platelets in solution (2 × 10⁷/ml) for 10 min before seeding onto immobilized surfaces at 37°C for 45 min followed by washing with PBS. Adherent platelets were fixed with 4% paraformaldehyde (PFA) at room temperature for 10 min before mounting on glass slides with Fluoromount G (Southern Biotech). Platelets were imaged using Kohler-illuminated Nomarski differential interference contrast optics with a Zeiss ×63 oil immersion 1.40 numerical aperture (NA) plan-apochromat lens on a Zeiss Axio Imager M2 microscope using Slidebook 5.5 image acquisition software (Intelligent Imaging Innovations, Denver, CO) as previously described (7).

Fluorescence microscopy.

After platelet spreading and PFA fixation as above, adherent platelets were permeabilized with blocking solution (1% BSA and 0.1% SDS in PBS). Platelets were then stained with indicated primary antibodies overnight at 4°C at a 1:100 dilution in blocking buffer. Alexa Fluor secondary antibodies (1:500) or TRITC-phalloidin (1:500) were added in blocking buffer for 2 h. Coverslips were mounted with Fluoromount G on glass slides. Platelets were imaged using a Zeiss Axio Imager M2 microscope. Adherent platelets were also imaged using superresolution structured illumination microscopy (SR-SIM) with a Zeiss ×100 oil immersion 1.46 NA lens on a Zeiss Elyra PS.1 microscope as previously described (8). For two-channel colocalization analyses, immunofluorescence overlap was quantified using an in-house Matlab application to calculate the Pearson’s coefficient as previously described (33). Data are shown as means ± SE. Statistical analyses were performed using the two-tailed Student’s t-test; P values < 0.05 were considered significant.

Immunoprecipitation, Ly-GDI protein capture, and Western blotting.

Purified platelets (0.5–1 × 10⁹/ml) were incubated with the glycoprotein (GP) IIbIIIa inhibitor Integrilin (20 µg/ml) before the addition of vehicle (0.1% DMSO) or Ro 31-8220 for 10 min at room temperature. Platelets were then stimulated with CRP (10 µg/ml) for 10 min at room temperature before the addition of lysis/immunoprecipitation (IP) buffer [10 mM Tris·HCl pH 7.4, 150 mM NaCl, 2 mM EDTA, 1% (vol/vol) Triton X-100] supplemented with phenylmethylsulfonyl (PMSF) and Sigma Phosphatase Inhibitor Cocktail 3. Platelet lysates were precleared with protein A/G Sepharose and then incubated with 2 μg of RhoGDI, D4-GDI antibodies, or nonspecific IgGs overnight at 4°C. Antibody protein complexes were then captured with protein A/G PLUS-agarose beads (2 h, room temperature) and washed three times in IP buffer. RhoGDI and Ly-GDI precipitates were then eluted through the addition Laemmli sample buffer (Bio-Rad, Hercules, CA) containing 200 mM DTT. Rac1–glutathione S-transferase (GST) protein capture assays were performed as previously described (5). Protein samples were resolved by SDS-PAGE and transferred to polyvinylidene difluoride membranes and blotted with indicated antibodies and horseradish peroxidase-conjugated secondary antibodies. Protein was detected using enhanced chemiluminescence (Thermo Scientific).

Chariot antibody delivery.

Chariot antibody interference experiments were performed following the methods of Diagouraga et al. (21) with minor modifications. Briefly, 2 μg of RhoGDI, Ly-GDI (D4-GDI) primary antibodies, or control IgGs were incubated with Chariot agent (2 mg/ml stock solution, diluted 1:25 with sterile H₂O before use) at room temperature for 45 min. Platelets (6 × 10⁷/ml) were added to the Chariot antibody solution at equal volume. The platelet-Chariot-antibody solution was then incubated at 37°C for 1 h. Samples were diluted with modified Tyrode buffer at equal volume before seeding onto fibrinogen-coated glass coverslips at 37°C for 45 min. Coverslips were then washed with PBS, fixed with 4% PFA, mounted, and visualized by differential interference contrast microscopy to measure and analyze platelet surface areas and adhesion as previously described (8).

In silico analyses.

High-probability (low false discovery rate) predicted Ly-GDI (ARHGDIB) protein-protein interactions were determined through FpClass queries (37) for molecules with a non-zero feature combination score and visualized as an interactome by GeNets (see https://apps.broadinstitute.org/genets). ChiBE (14) was used to query and visualize the literature-curated interaction neighborhood of Ly-GDI (ARHGDIB) in Pathway Commons (18) and Ly-GDI phosphorylations in PhosphoSitePlus (32). The “controls-state-change-of” binary relation type to capture signaling and “in-complex-with” and “interacts-with” relation types were used to capture protein interactions, and these two interaction types were merged for simpler visualization.

RESULTS

Expression of RhoGDI and Ly-GDI in human platelets.

The Rho GTPases Rac1, Cdc42, and RhoA have several reported roles in platelet function, yet whether they are associated with or regulated by Rho-specific guanine nucleotide dissociation inhibitor (RhoGDI) proteins in platelets remains unexamined (10). We first sought to determine whether platelets express RhoGDI family members, including RhoGDI and Ly-GDI. Washed human platelets were prepared from citrated whole blood from healthy donors and lysed directly into Laemmli sample buffer before SDS-PAGE and Western blot analysis for RhoGDI and Ly-GDI expression. Whole cell lysates of MDA-MB-231 cells, a breast cancer cell line that expresses functional RhoGDI as well as Ly-GDI (56), served as a positive control for GDI protein expression and immunoreactivity. As seen in Fig. 1A, Western blotting determined that human platelets express RhoGDI as well as Ly-GDI at levels similar to MDA-MB-231 cells relative to the expression of actin, which served as a loading control.

Fig. 1.

Expression and localization of RhoGDI proteins in human platelets. A: Western blot analysis of RhoGDI and Ly-GDI expression in human platelet (5 μl, 1 × 10⁹ platelets/ml, 1:1 Laemmli sample buffer) and MDA-MB-231 cell lysates. Actin serves as a loading control. Positions of 28-kDa and 48-kDa molecular mass markers are indicated with tick marks as shown. B: replicate samples of washed human platelets (2 × 10⁷/ml) were spread on glass coverslips coated with fibrinogen before fixation, blocking, and staining for RhoGDI or Ly-GDI (green) together with actin (red, stained with TRITC-phalloidin) and visualization by immunofluorescence microscopy. Scale bar, 10 μm. Representative platelets (white box) from each set of stainings are also shown at an enhanced ×3.5  magnification.

RhoGDI and Ly-GDI specifically distribute in platelets.

To examine the expression and intracellular localization of RhoGDI and Ly-GDI proteins in platelets, we next performed immunofluorescence microscopy analyses of platelets adherent to a surface of fibrinogen. Washed human platelets were incubated on fibrinogen-coated coverglass (37°C, 45 min) before fixation, blocking and staining with RhoGDI or Ly-GDI antibodies and TRITC-phalloidin to visualize the platelet actin cytoskeleton. In platelets adherent to fibrinogen, RhoGDI showed a distributed, granular-type staining pattern, while Ly-GDI displayed an uneven or “polarized” localization, often adjacent to or emanating from the centralized platelet granulomere at nascent lamellipodia-like structures (Fig. 1B).

Rac1 and Cdc42 colocalize with Ly-GDI in platelets.

As GDI proteins bind to Rho GTPases to sequester and release their activities at specific membrane and intracellular compartments (25, 31), we next examined the intracellular localization of RhoGDI or Ly-GDI together with the Rho GTPases Rac1, Cdc42, or RhoA in fibrinogen-bound platelets by immunofluorescence superresolution structured illumination microscopy (SR-SIM). Cdc42 and Rac1 showed minimal overlap with RhoGDI in platelets spread on fibrinogen (Fig. 2A). Interestingly, Rac1 and Cdc42 both showed a polarized distribution largely overlapping with Ly-GDI and suggestive of structural elements (Fig. 2B) resembling the localization of microtubules in platelets adherent to fibrinogen, as previously described (15, 34, 36, 48, 50, 52). Unlike Cdc42 and Rac1, however, RhoA showed a more evenly distributed staining pattern throughout the platelet cell body that did not overlap with RhoGDI or Ly-GDI (Fig. 2, A–C). Pearson’s correlation-based colocalization analyses revealed more colocalization between Ly-GDI and Rac1 as well as Ly-GDI and Cdc42 (Fig. 2D) relative to RhoGDI and the Rho GTPases (Fig. 2D). Pearson’s analyses also supported a greater extent of colocalization between Rac1 and GDI proteins in general compared with Cdc42 in a manner reflective of the qualitatively more structured organization of Cdc42 in adherent platelets (Fig. 2, C and D). As a functional control to confirm that platelet Ly-GDI can interact with Rho GTPases, we captured Ly-GDI protein from platelet lysates with immobilized Rac1-GST protein. As seen in Fig. 2E, GST-Rac1 captured Ly-GDI from resting as well as CRP-stimulated platelet lysates at equivalent levels, consistent with the role of Ly-GDI in Rho GTPase sequestration. GST protein alone failed to capture any detectable Ly-GDI protein from platelet lysates (Fig. 2E).

Fig. 2.

Localization of RhoGDI and Ly-GDI with Rac1 and Cdc42 in platelets adherent to fibrinogen. A and B: replicate samples of washed human platelets (2 × 10⁷/ml) were spread on glass coverslips coated with fibrinogen before fixation, blocking, and staining for RhoGDI (A) or Ly-GDI (green) (B) together with Rac1, Cdc42, or RhoA (red) and visualization by SR-SIM. Scale bar, 2 μm. C: magnified images of individual platelets representative of the colocalization Ly-GDI (green) and RhoA, Rac1, or Cdc42 (red) SR-SIM. Scale bar, 2 μm. D: Pearson’s correlation of GDI and Rho GTPase overlay demonstrates significantly increased colocalization of Ly-GDI and Rac1 or Cdc42 relative to RhoGDI; determined by Student’s t-test, and *P < 0.05. E: Ly-GDI protein capture by Rac1-GST and GST alone from resting and CRP-activated platelet lysates analyzed by Western blotting (WB). For WB panels, tick marks indicate position of 28-kDa molecular mass marker. GST protein loading for protein capture shown by PAGE and Coomassie (Coom.) staining. For the Comassie gel, the tick marks indicate position of 28 kDa (bottom), 35-kDa (middle), and 48-kDa (top) molecular mass markers.

Ly-GDI colocalizes with polarized cytoskeletal elements in platelets.

Recent models of platelet function suggest a polarized role for Rho GTPase activities in platelet attachment, spreading, and secretion during hemostatic plug or thrombus formation (45, 49, 54), but cell biological mechanisms to account for polarized Rho GTPase activities have not yet been set forth. Intriguingly, the microtubules of adherent platelets, originating from a specific unwinding and reorganization of the marginal band during activation (11, 21, 46, 47), often show a shifted or polarized distribution as platelets spread on surfaces such as fibrinogen (15). We therefore next examined whether Ly-GDI colocalized with microtubules in platelets. Platelets adherent to fibrinogen were stained for Ly-GDI and α-tubulin and processed for immunofluorescence microscopy. As seen in Fig. 3A, the microtubules in platelets adherent to fibrinogen displayed a partially unwound structure shifted in a polarized-like manner. Costaining experiments demonstrated a colocalization of these polarized microtubules together with Ly-GDI (Fig. 3A). Superresolution microscopy studies further supported a specific colocalization of Ly-GDI together with specific microtubule structures (Fig. 3B). To further investigate the punctate, granular-like staining pattern of RhoGDI, we next examined the colocalization of RhoGDI or Ly-GDI together with the α-granule marker P-selectin. As seen in Fig. 3C, Ly-GDI staining showed relatively more overlap with P-selectin compared with RhoGDI. RhoGDI and Ly-GDI showed minimal colocalization with the platelet integrin β₃ (Fig. 3D).

Fig. 3.

Localization of RhoGDIs with structural elements in platelets adherent to fibrinogen. A: replicate samples of washed human platelets (2 × 10⁷/ml) were spread on glass coverslips coated with fibrinogen before fixation, blocking, and staining for RhoGDI or Ly-GDI (green) and α-tubulin (red) and visualization by SR-SIM fluorescence microscopy (scale bar, 2 μm). Pearson’s correlation reveals significantly increased colocalization of Ly-GDI and α-tubulin relative to RhoGDI and α-tubulin (determined by Student’s t-test; indicated by *). B: magnified images of individual platelets representative of the colocalization RhoGDI or Ly-GDI (green) and α-tubulin (scale bar, 2 μm). C and D: other sets of coverslips were stained for RhoGDI or Ly-GDI (green) together with the platelet α-granule marker P-selectin (red) (C) and integrin β3 (red) (D) and visualized by conventional fluorescence microscopy. Scale bar, 10 μm.

Ly-GDI interference inhibits platelet spreading.

The localization of Ly-GDI together with Rac1 and Cdc42 at polarized cytoskeletal elements suggested a role for Ly-GDI in the sequestration, localization, and regulation of Rho GTPase function in platelets. Accordingly, we next sought to examine whether interfering with Ly-GDI or RhoGDI would impact platelet function. To interfere with RhoGDI accessibility and any putative Rho GTPase–associated functions in platelets, a Chariot agent permeabilization protocol (21) was used to introduce antibodies that bind to the NH₂ terminal, Rho GTPase binding domains of RhoGDI or Ly-GDI into live, washed human platelets before spreading on fibrinogen-coated coverslips. As seen in Fig. 4, A and B, Chariot-permeabilized platelets treated with nonspecific IgGs or RhoGDI antibodies spread on fibrinogen normally with mean surface areas of 20.7 ± 0.23 μm² and 22.7 ± 1.69 μm², respectively; in contrast, treatment of platelets with anti-Ly-GDI antibodies dramatically reduced spreading on fibrinogen (9.48 ± 0.71 μm²). Chariot-permeabilized platelets treated with Ly-GDI antisera also exhibited a reduced binding to fibrinogen (24.7 ± 2.91 bound platelets per field vs. 69.7 ± 15.5 and 65.3 ± 11.7 bound platelets per field for IgG and anti-RhoGDI-treated platelets, respectively).

Fig. 4.

Ly-GDI interference inhibits platelet spreading on fibrinogen. Antibodies directed against the NH₂-terminal Rho GTPase binding domains of RhoGDI or Ly-GDI or control IgGs were introduced into washed human platelets using the Chariot system. A–C: platelets were spread on fibrinogen-coated coverglass for 45 min before fixation, mounting, and visualization by differential interference contrast Nomarski microscopy (A) to assess the statistical distribution of surface areas of platelets adherent to fibrinogen (B) and to quantify the numbers of adherent platelets per field (C). Scale bar, 10 μm.

Ly-GDI phosphorylation in platelet activation.

A number of signaling proteins such as the Src family tyrosine kinases and the p21-activated kinases serve regulatory roles in platelets upstream and downstream of Rac1-GTP and Cdc42-GTP formation, respectively (10, 31). However, in platelets, Rac1 and Cdc42 themselves are not typically regulated by phosphorylation, but rely on the phosphorylation of regulatory proteins such as GEFs and GAPs to translate intracellular signals into Rho GTPase functional outputs. To begin to investigate and visualize how Ly-GDI may regulate platelet function mechanistically, we next examined Ly-GDI protein-protein interactions and functional covalent modifications using both predicative and curated bioinformatics tools. An in silico analysis of Ly-GDI (termed “ARHGDIB”) binding partners based on modeled, folded domain structures using FpClass predicted 19 high-probability protein-protein interactions with Rac1, Cdc42, and RhoA as well as several other Rho GTPases and Rho GTPase regulators such as Vav1 (Fig. 5A). These and other functional interactions were additionally evident in curated protein interaction data in Pathway Commons, which also suggested roles for Caspase-3 (Fig. 5B) as well as specific protein kinases (PKCα, Src) in Ly-GDI regulation (Fig. 5, B and C).

Fig. 5.

Ly-GDI is phosphorylated at PKC substrate motifs following platelet activation and colocalizes with PKC in adherent platelets. A: FpClass predicted interaction partners for Ly-GDI (ARHGDIB) visualized by GeNets. B: Pathway Commons neighborhoods of curated Ly-GDI (ARHGDIB) interaction partners and regulatory proteins. Directed edges (arrows) indicate signaling steps and undirected edges indicate protein interactions. C: ChiBE “detailed view” of the specific phosphorylations of Ly-GDI by PKCα (PRKCA) as well as Src and Syk as curated by PhosphoSitePlus. D: washed human platelets (1 × 10⁹/ml) were stimulated with CRP or vehicle alone (10 min, 37°C) before collection into IP buffer and immunocapture with Ly-GDI antibodies or nonspecific rabbit IgGs. After IP, protein A/G eluates were analyzed for PKC substrate phosphorylation and total Ly-GDI protein capture by Western blotting (WB). For WB panels, tick marks indicate position of 28-kDa molecular mass marker. E: replicate samples of washed human platelets (2 × 10⁷/ml) were spread on glass coverslips coated with fibrinogen before fixation and staining for RhoGDI or Ly-GDI (green) together with PKC (red) and visualized by conventional fluorescence microscopy (scale bar, 10 μm). Pearson’s correlation reveals significantly increased colocalization of Ly-GDI and PKC relative to RhoGDI and PKC (indicated by *). F: wide-field (scale bar, 5 μm) and magnified (scale bar, 2 μm) superresolution structured illumination microscopy imaging of adherent platelets stained for Ly-GDI (green) and PKC (red).

To investigate whether RhoGDI and Ly-GDI serve as substrates of these kinases with known roles in Rac1 and Cdc42 regulation in the platelet activation program, we next examined the phosphorylation status of RhoGDI and Ly-GDI in resting and CRP-stimulated platelets by Western blot analysis following Ly-GDI immunoprecipitation. Washed human platelets were treated with the GPVI-specific agonist CRP for 10 min before lysis into IP buffer and immunocapture with RhoGDI, Ly-GDI or control IgG antisera. After protein A/G agarose precipitation, Western blot analysis revealed that Ly-GDI was readily captured from platelet lysates (Fig. 5D). Western blotting with antisera directed against phosphorylated consensus protein kinase C (PKC) (R/KXpSX[R/K]) substrate motifs determined that Ly-GDI was phosphorylated at PKC motifs following stimulation with CRP. Although Ly-GDI has been previously reported to be regulated by tyrosine phosphorylation in other cellular systems to influence physiological processes such as cancer cell metastasis (55), no tyrosine phosphorylation of platelet Ly-GDI was detectable by IP and Western blot analysis with 4G10 antisera under the conditions of our study (data not shown). Similarly, we did not detect any caspase-mediated proteolysis of Ly-GDI in CRP-activated platelets using antisera that specifically recognize the cleaved form of Ly-GDI (data not shown). While RhoGDI readily precipitated from platelet lysates, no phosphorylation of the platelet RhoGDI protein was detected by Western blot analysis under the conditions described (data not shown).

The detected phosphorylation of Ly-GDI at PKC motifs raised the possibility that PKC may also localize with Ly-GDI to regulate platelet function. Accordingly, we next examined the localization of Ly-GDI together with PKC in platelets adherent to fibrinogen. While PKC showed some colocalization with RhoGDI, Ly-GDI colocalized more significantly with PKC in adherent platelets in a polarized manner as determined by fluorescence microscopy (Fig. 5E), SR-SIM (Fig. 5F), and Pearson’s correlation analyses.

PKC regulates Ly-GDI phosphorylation and distribution in platelets.

The colocalization of Ly-GDI with PKC suggested a role for PKC in regulating Ly-GDI function in the coordination of platelet Rho GTPase activities. To further investigate the regulation of Ly-GDI by PKC, we next examined the phosphorylation of Ly-GDI under PKC-inhibited conditions. Purified platelets were pretreated the pan-PKC inhibitor Ro 31-8220 and the PKCαβ-specific inhibitor Go 6976 or vehicle alone (0.1% DMSO) for 10 min before stimulation with CRP (10 μg/ml, 10 min), lysis into IP buffer, and immunocapture with Ly-GDI or control IgG antisera. Western blot analysis of precipitates revealed that Ly-GDI was readily enriched from platelet lysates under all treatment conditions (Fig. 6, A and B) and that PKC substrate phosphorylation of Ly-GDI increased following CRP treatment, but not under PKC-inhibited conditions.

Fig. 6.

PKC inhibition blocks Ly-GDI phosphorylation and disorganizes Ly-GDI localization in adherent platelets. A and B: replicate samples of purified human platelets (1 × 10⁹/ml) were pretreated with the PKC inhibitors Ro 31-8220 (10 μM) or Go 6976 or vehicle alone (0.1% DMSO) before stimulation with CRP and immunoprecipitation (IP) with Ly-GDI or nonspecific IgG antisera. Eluates were analyzed for PKC substrate phosphorylation and total Ly-GDI protein capture by Western blottomg (WB). For WB panels, tick marks indicate position of 28-kDa molecular mass marker. C: replicate samples of washed human platelets (2 × 10⁷/ml) were spread on glass coverslips coated with fibrinogen before fixation and staining for Ly-GDI (green) together with actin (red) in the presence of Ro 31-8220 (10 μM), Go 6976 (1 μM), or vehicle alone (0.1% DMSO). Scale bar, 5 μm. Pearson’s correlation of Ly-GDI and actin colocalization serves as a measure of the loss of Ly-GDI polarization. Significance (*P < 0.05) determined by Student’s t-test. D: other sets of coverslips were stained for Ly-GDI (green) together with Rac1 (red) or Cdc42 (red) as visualized by SR-SIM (scale bar, 5 μm). E: representative, magnified SR-SIM visualization of Ly-GDI (green) and Rac1 (red) colocalization and distribution in platelets adherent to fibrinogen (scale bar, 2 μm).

To examine whether PKC influences the intracellular, polarized distribution of Ly-GDI in platelets, we visualized Ly-GDI together with actin in platelets adherent to fibrinogen under control and PKC-inhibited conditions. As seen in Fig. 6C, compared with vehicle alone, Ro 31-8220 and Go 6976 treatment partially inhibited platelet spreading on fibrinogen and disorganized the polarized, concentrated localization of Ly-GDI in adherent platelets, as visualized by SR-SIM and measured by the increased Pearson’s correlation of Ly-GDI overlap with actin. However, PKC inhibition had no significant effect on the distribution of the Rho GTPases Rac1 and Cdc42, as also determined by SR-SIM and Pearson’s correlation of Rho GTPase and Ly-GDI colocalization (Fig. 6D). While Ly-GDI showed a less structured and more diffuse distribution in platelets spread on fibrinogen in the presence of Ro 31-8220 and Go 6976 (Fig. 6, D and E), both Rac1 and Cdc42 remained polarized, suggesting a role for PKC in spatially coordinating regulatory interactions between Ly-GDI and Rac1 or Cdc42 in activating platelets.

DISCUSSION

Here we report that platelets express the Rho-specific guanine nucleotide dissociation inhibitor proteins RhoGDI and Ly-GDI and provide data supporting roles for Ly-GDI, a GDI family member commonly found in lymphocytic and hematopoietic cells, in platelet function. We found that while platelets express both RhoGDI and Ly-GDI proteins at similar levels, these GDI proteins show distinct intracellular granular and “polarized” cytoskeletal localizations, respectively, in platelets. Ly-GDI colocalized with the Rho GTPase proteins Rac1 and Cdc42 as well as α-tubulin, consistent with a role in regulating Rho GTPase activities at areas of dynamic cytoskeletal remodeling. Antibody interference experiments also supported a role for Ly-GDI in platelet spreading and Rho GTPase function, as antibodies targeting the NH₂ terminal, Rho GTPase binding domain of Ly-GDI introduced into live, purified platelets inhibited platelet spreading on fibrinogen. Biochemical experiments demonstrated that Ly-GDI is also a target of signaling systems in the platelet activation program, including PKCs, which may promote the phosphorylation of platelet Ly-GDI directly downstream of platelet GPVI engagement or secondarily in response to the release of agonists such as ADP following CRP stimulation. In addition to Rac1 and Cdc42, cytoskeletal-associated Ly-GDI also colocalized with PKC in adherent platelets, and we found that PKC inhibition abrogated Ly-GDI phosphorylation and polarization while leaving Rac1 and Cdc42 polarization intact. Together, our results support roles for Ly-GDI in platelet function and suggest that PKC may orchestrate the regulation of Rac1 and Cdc42 in platelets through Ly-GDI phosphorylation to spatially coordinate and fine tune Rho GTPase–associated processes during platelet activation (Fig. 7).

Fig. 7.

Model of hypothesized role of Ly-GDI in platelet function. After the engagement of platelet receptors (GPVI, integrin α_IIbβ₃, etc.), Src family kinases (SFKs) drive the protein kinase C (PKC)-mediated phosphorylation, activation, and localization of Ly-GDI to organize Rho GTPase (Rac1, Cdc42, etc.) activities in concert with Rho GTPase-activating proteins (GAPs) and guanine nucleotide exchange factors (GEFs) to direct cytoskeletal remodeling and platelet function.

In mammalian cells, over 20 different functionally diverse Rho GTPase proteins are regulated by ~60 different GAP and GEF proteins; yet only a total of three RhoGDI gene products serve to direct Rho GTPase activities intracellularly (30, 31). In platelets, three Rho GTPase family members, Rac1, Cdc42, and RhoA, play critical roles in platelet function. Moreover, recent studies have identified a role for Rho proteins such as RhoG in regulating platelet physiology (10, 26). While a number of signaling systems such as tyrosine kinases have been shown to regulate Rho GTPase activities in platelets and other cells, the Rho GTPases themselves are not directly regulated by phosphorylation in regards to platelet function. Accordingly, as is the case in other cell types, a number of Rho GTPase regulatory proteins most likely serve as targets of signaling pathways to mediate Rho GTPase activation, organization, and function in platelets. Notably, the Rho GEF Vav, a component of the linker for activation of T cells signalosome and target of early phosphorylation events in platelet activation, has been shown to play a role in regulating platelet Rac1 activation and platelet function (10, 40, 53). Other GEFs such as phosphatidylinositol 3,4,5-triphosphate-dependent Rac exchanger, leukemia-associated RhoGEF, TIAM, p21-activated kinase-interacting exchange factor, and GIT proteins have been investigated for roles in platelet function (10). Studies of GAPs such as IQGAP, ARHGAP17, Nadrin, and oligophrenin-1 have described functions for these proteins in platelets. Yet, despite their interactions with Rho GTPases and a number of other proteins and signaling systems critical to platelet function (1, 10, 25, 29, 55), RhoGDIs have remained largely overlooked in studies of platelet Rho GTPase regulation. Here we find that platelets express RhoGDI and Ly-GDI and provide data suggesting specific roles for Ly-GDI in platelet function.

In an effort to understand how platelet RhoGDI proteins may be regulated by intracellular signaling systems associated with platelet activation programs, we used a set of bioinformatics tools to determine predicted and known GDI binding partners and modifying enzymes (Fig. 5, A–C). PhosphoSitePlus-based Pathways Commons analyses of Ly-GDI regulation noted the regulation of Ly-GDI by PKC. In vitro IP-Western blot analyses demonstrated that Ly-GDI is phosphorylated at PKC consensus sites in activating platelets and that Ly-GDI colocalized with PKC as well as Rac1 and Cdc42 in spreading platelets, potentially linking PKC signaling to platelet Rho GTPase regulation (Fig. 5). In other cell types a number of regulatory mechanisms are known to converge on RhoGDIs to effect cellular outputs. For instance, prenylation serves to link RhoGDIs to specific membranes. 14-3-3 protein binding and release has also been proposed to regulate RhoGDI localization and function. Ly-GDI has also been shown to be regulated by proteolytic cleavage (23). Reversible lysine acetylation as catalyzed by lysine acetyltransferases such as p300 and deacetylase enzymes such as histone deacetylase 6 (HDAC6) and sirtuin (SIRT) proteins have also been shown to modify RhoGDIs to regulate their function (38). Indeed, our recent characterization of the platelet lysine acetylome identified Ly-GDI as a lysine-acetylated protein in platelets with a putative role in regulating actin-mediated platelet processes (6, 12). Future work is needed to better understand how the phosphorylation and lysine acetylation of Ly-GDI as well as specific Ly-GDI protein-protein interactions regulate platelet function.

Present models of Rho GTPases in platelet cell biological functions account for a temporal sequence of platelet activation events, specifically attachment, secretion, spreading—generally medicated by Rac1 and Cdc42—and early and late contractile events such as shape change and retraction as mediated by RhoA–Rho-associated protein kinase signaling (4, 10). Although the global, temporal regulation of these events in platelets has been investigated over the past decade, more recent studies suggest that platelet activation is spatially regulated to solicit specific secretory and cytoskeletal outputs within a tightly packed “core” or more loosely associated expanding “shell” of a growing hemostatic plug or thrombus (49, 54). Along these lines, hypotheses of differential Rho GTPase activation within a growing hemostatic plug or thrombus are supported by studies demonstrating that the spatial context of adhesive matrix substrates translates into specific, organized Rac1-dependent intracellular cytoskeletal and secretory platelet responses (45, 49). However, mechanisms regarding how the polarization of such signaling events may come about in platelets have not been proposed. Intriguingly, we find that Ly-GDI localizes in a polarized manner in adherent platelets, in an intracellular area overlapping with microtubules, Rac1, Cdc42, and PKC, proposing that Ly-GDI acts as a putative, spatially defined regulator of Rac1 and Cdc42 activities in platelets. Together with studies showing an organized cross talk of the platelet actin and microtubule cytoskeletons (15, 19, 46, 50), our findings provide insights into how signaling events may be spatially organized relative to the dynamic cytoskeleton of activating platelets through Ly-GDI in the physiological context of hemostatic plug and thrombus formation (Fig. 7).

GRANTS

Superresolution microscopy studies were supported in part by the M. J. Murdock Charitable Trust. This work was supported by the Knight Cardiovascular Institute and grants from the National Institutes of Health (CA143971 to D. Theodorescu and R01 HL-101972 and R01 GM-116184 to O.J.T. McCarty) and the American Heart Association (17SDG33350075 to J. E. Aslan, 13EIA12630000 to O.J.T. McCarty, and 16UFEL31530001 to M. Thierheimer. M. Thierheimer is an OSU Johnson Scholar.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

A.T.N. and J.E.A. conceived and designed the research; A.T.N., M.L.T., J.P., R.A.R., A.M., and J.E.A. performed experiments; A.T.N., Ö.B., A.D.R., T.H., J.B., X.N., and J.E.A. analyzed data; A.T.N., Ö.B., A.D.R., T.H., X.N., O.J.T.M., and J.E.A. interpreted results of experiments; A.T.N., Ö.B., A.D.R., J.B., and J.E.A. prepared figures; A.T.N. and J.E.A. drafted manuscript; A.T.N., Ö.B., R.A.R., D.T., J.B., E.D., O.J.T.M., and J.E.A. edited and revised manuscript; A.T.N., M.L.T., Ö.B., A.D.R., T.H., J.P., R.A.R., A.M., D.T., J.B., X.N., E.D., O.J.T.M., and J.E.A. approved final version of manuscript.