DOCK8 Protein Regulates Macrophage Migration through Cdc42 Protein Activation and LRAP35a Protein Interaction*

DOCK8 is an atypical guanine nucleotide exchange factor for Cdc42, and its mutations cause combined immunodeficiency in humans. Accumulating evidence indicates that DOCK8 regulates the migration and activation of various subsets of leukocytes, but its regulatory mechanism is poorly understood. We here report that DOCK8-deficient macrophages exhibit a migration defect in a 2D setting. Although DOCK8 deficiency in macrophages did not affect the global Cdc42 activation induced by chemokine stimulation, rescue experiments revealed that the guanine nucleotide exchange factor activity of DOCK8 was required for macrophage migration. We found that DOCK8 associated with LRAP35a, an adaptor molecule that binds to the Cdc42 effector myotonic dystrophy kinase-related Cdc42-binding kinase, and facilitated its activity to phosphorylate myosin II regulatory light chain. When this interaction was disrupted in WT macrophages, they showed a migration defect, as seen in DOCK8-deficient macrophages. These results suggest that, during macrophage migration, DOCK8 links Cdc42 activation to actomyosin dynamics through the association with LRAP35a.

Cdc42 is a member of the small GTPases that function as molecule switches by cycling between GDP-bound inactive and GTP-bound active states (1). The stimulus-induced formation of active Cdc42 is mediated by guanine nucleotide exchange factors (GEFs), 2 and, when activated, Cdc42 binds to multiple effector molecules (1). One such effector molecule is myotonic dystrophy kinase-related Cdc42-binding kinase (MRCK), which stimulates the phosphorylation of myosin II regulatory light chain (MLC2) and controls actomyosin dynamics (2)(3)(4)(5). Cdc42 is known to act as a master regulator of cell polarity in eukaryotic organisms ranging from yeasts to humans (1). In addition, it has been shown that Cdc42-deficient dendritic cells (DCs), a subset of antigen-presenting leukocytes, fail to migrate effectively within 3D extracellular matrix scaffolds, whereas they exhibit only limited defects in a 2D setting (6). Thus, Cdc42 plays a unique role in leukocyte migration.
DOCK8 is a member of the evolutionarily conserved DOCK family proteins that function as GEFs for the Rho family of GTPases (7,8). Although DOCK8 does not contain the Dbl homology domain typically found in GEFs, it does bind to Cdc42 and mediates the GTP-GDP exchange reaction through the DOCK homology region 2 (DHR2) domain (9). Recently, substantial attention has been paid to the signaling and functions of DOCK8, as it has been reported that Dock8 mutations cause combined immunodeficiency syndrome in humans, with morphological and functional abnormalities of leukocytes (10 -15). The role of DOCK8 in leukocytes was also demonstrated using animal models. For example, the N-ethyl-N-nitrosoureamediated mutagenesis in mice has shown that DOCK8 plays an important role in B cell immunological synapse and long-lasting humoral immunity and also contributes to the development or survival of memory CD8 ϩ T cells (15)(16)(17). Furthermore, by generating DOCK8-deficient (Dock8 Ϫ/Ϫ ) mice, we and others have revealed that DOCK8 is required for DCs to migrate effectively within 3D, but not 2D, microenvironments (9,18). However, DOCK8 deficiency in DCs did not affect the global Cdc42 activation induced by chemokine stimulation (9). Therefore, how DOCK8 regulates leukocyte migration remains unclear.
Macrophages are a subset of leukocytes involved in innate immune responses, and they play important roles in the induction of protective immunity to invading pathogens (19). We found that, even in a 2D setting, Dock8 Ϫ/Ϫ macrophages exhibited reduced motility with a defect in nuclear positioning. Although DOCK8 deficiency in macrophages also did not affect global Cdc42 activation, rescue experiments revealed that the GEF activity of DOCK8 was required for macrophage migration. In this study, we show that DOCK8 associates with LRAP35a, an adaptor molecule that binds to MRCK (4), and facilitates MLC2 phosphorylation. Our findings suggest that, during leukocyte migration, DOCK8 links Cdc42 activation to actomyosin dynamics through its association with LRAP35a. * This work was supported by a grant-in-aid for scientific research on innovative areas from the Ministry of Education, Culture, Sports, Science, and Technology of Japan; grants-in-aid for scientific research from the Japan Society for the Promotion

DOCK8 Regulates Macrophage Migration in 2D
Environments-To examine the role of DOCK8 in macrophages, we first compared the recruitment of macrophages to inflamed lungs between Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ littermate mice. In response to intratracheal LPS challenge, macrophages rapidly accumulated in the alveolar cavity of Dock8 ϩ/Ϫ mice (Fig. 1A). However, such accumulation was impaired in Dock8 Ϫ/Ϫ mice (Fig.  1A). It is known that this recruitment of macrophages critically depends on the chemokine CCL2 in alveolar tissues (20). To compare the chemotactic response to CCL2 in vitro, we differentiated macrophages from Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ bone marrow (BM) cells in the presence of M-CSF. Irrespective of DOCK8 expression, these BM-derived macrophages exhibited an adherent morphology (Fig. 1B), and the majority were CD11c low CD11b high cells (Fig. 1C). These phenotypes were different from the DCs generated from BM cells with GM-CSF (Fig. 1C). When Dock8 ϩ/Ϫ macrophages were stimulated with CCL2 in an EZ-TAXIScan chamber, their leading edges efficiently moved toward the chemokine source at a speed of 1.77 m/min (Fig. 1D). Although DOCK8 deficiency did not affect the directionality and y-forward migration index during macrophage migration (Fig. 1E), the average speed of Dock8 Ϫ/Ϫ BMderived macrophages was reduced to 61.0% of the Dock8 ϩ/Ϫ speed (Fig. 1D). Similar results were obtained when thioglycollate (TG)-elicited peritoneal macrophages were isolated from Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ mice and analyzed for migration speed (1.11 versus 0.61 m/min, Fig. 1F). However, no such defect was observed with Dock8 Ϫ/Ϫ BM-derived DCs in this 2D setting (Fig. 1G), as reported previously (9).
During the in vitro chemotaxis assay, we found that nuclear positioning was different between Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ macrophages. Both Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ BM-derived macrophages undergoing chemotaxis actively formed ring-like structures at the leading lamella ( Fig. 1B and supplemental Movie 1). This was more clearly observed when cells were stained with phalloidin ( Fig. 1H). However, although the nucleus was located immediately below the ring-like structure in Dock8 ϩ/Ϫ BM-derived macrophages, such localization of the nucleus was hardly detected in Dock8 Ϫ/Ϫ macrophages ( Fig. 1, B and H, and supplemental Movie 1). Indeed, when the relative positions of the nucleus within the cell (the distance from the nucleus to the rear per whole cell length) were compared 60 min after migration, the values of Dock8 Ϫ/Ϫ BM-derived macrophages were significantly lower than those of Dock8 ϩ/Ϫ BM-derived macrophages (0.68 versus 0.44) (Fig. 1I). Similar results were obtained when TG-elicited peritoneal macrophages were analyzed (0.62 versus 0.47, Fig. 1J).
DOCK8 Regulates Macrophage Migration through Cdc42 Activation-Having found that DOCK8 regulates macrophage migration, we next examined whether the GEF activity of DOCK8 is required for this process. By analogy of the DOCK9 DHR2 domain, which is also known to act as a Cdc42 GEF (21), the valine residue at position 1986 of DOCK8 was expected to function as a nucleotide sensor. Indeed, biochemical analyses revealed that the Cdc42 GEF activity of the recombinant DOCK8 DHR2 protein was completely abolished by mutating this valine residue to alanine (VA mutant) despite the quality of the purified proteins being unaffected. (Fig. 2, A-C). When the GFP-tagged WT and VA mutant were expressed in Dock8 Ϫ/Ϫ BM-derived macrophages using Amaxa Nucleofector TM II, they showed comparable expression (Fig. 2D). However, although the migration speed was restored in Dock8 Ϫ/Ϫ BMderived macrophages expressing WT DOCK8 (Fig. 2E), the expression of the GEF-dead VA mutant failed to rescue the migration defect (Fig. 2E). These results indicate that DOCK8 regulates macrophage migration through Cdc42 activation. However, as seen in Dock8 Ϫ/Ϫ DCs (9), DOCK8 deficiency in BM-derived macrophages did not affect the global Cdc42 activation induced by CCL2 stimulation (Fig. 2F), probably because of functional compensation by other Cdc42 GEFs.
DOCK8 Regulates Actomyosin Dynamics at the Leading Lamella during Macrophage Migration-Nonmuscle myosin II (NMII) is composed of two heavy chains, two essential light chains, and two regulatory light chains (22). Although it is well known that NMII mediates retraction of the rear during cell migration (22,23), several studies have indicated that NMII also regulates leading edge motion (24 -27). To examine the localization of NMII, we stained Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ BM-derived macrophages with antibody for NMIIA. Although the staining pattern of NMIIA at the cell rear was unchanged between Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ BM-derived macrophages, the localization and accumulation of NMIIA at the leading lamella were less evident in Dock8 Ϫ/Ϫ BM-derived macrophages (Fig. 3, A and B). When GFP-tagged MLC2 was retrovirally expressed in Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ BM-derived macrophages (Fig. 3C), the migration speed was reduced, but a difference still existed between Dock8 ϩ/Ϫ and Dock8 Ϫ/Ϫ macrophages (1.43 versus 0.92 m/min, Fig. 3D). More importantly, the dynamic movement of MLC2 was found at the leading lamella of Dock8 ϩ/Ϫ but not Dock8 Ϫ/Ϫ BM-derived macrophages (Fig. 3C). The dynamics of actomyosin critically depend on the phosphorylation status of MLC2 (22,23). In response to CCL2 stimulation, MLC2 was phosphorylated at the serine residue of position  in Dock8 ϩ/Ϫ BM-derived macrophages (Fig. 3E). Immunofluorescence staining revealed that phosphorylated MLC2 was localized at the leading lamella and subnuclear region (Fig. 3F). However, such phosphorylation was significantly reduced in Dock8 Ϫ/Ϫ macrophages (Fig. 3, E and F). These results suggest that DOCK8 regulates the actomyosin dynamics during macrophage migration through MLC2 phosphorylation.
DOCK8 Links Cdc42 Activation to Actomyosin Dynamics through LRAP35a Interaction-MRCK␣ is a Cdc42 effector that mediates MLC2 phosphorylation at Ser-19 in collaboration with the leucine-rich adaptor protein LRAP35a (2)(3)(4)(5). During the search for binding partners of DOCK8, we found that DOCK8 and LRAP35a were coimmunoprecipitated when expressed in HEK-293T cells (Fig. 4A). Similar results were obtained when cell extracts of BM-derived macrophages were pulled down with GST fusion LRAP35a (Fig. 4B). On the other hand, ␤PIX (28), another Cdc42 GEF, did not show any binding to LRAP35a (Fig. 4A). A previous report indicated that the N-terminal region of LRAP35a containing leucine repeats is important for MRCK binding (4). In contrast, pulldown experiments using recombinant proteins revealed that LRAP35a associates with DOCK8 through the C-terminal region (Fig.  4C). When LRAP35a expression was knocked down in Dock8 ϩ/Ϫ BM-derived macrophages using siRNA (Fig. 4D), the migration speed was reduced to the level of Dock8 Ϫ/Ϫ macrophages (Fig. 4E). Similar results were obtained by overexpressing an LRAP35a N mutant that failed to bind to DOCK8 (Fig.  4F). In addition, knockdown of either LRAP35a or MRCK␣ suppressed MLC2 phosphorylation at Ser-19 (Fig. 4, G and H). These results suggest that, during macrophage migration, DOCK8 acts as a signaling molecule to integrate Cdc42 activation and actomyosin dynamics through its interaction with LRAP35a.
Role of LRAP35a in DC Migration in 2D and 3D Environments-Although Dock8 Ϫ/Ϫ BM-derived DCs migrate normally on 2D surfaces, they fail to migrate effectively within 3D extracellular matrix scaffolds such as collagen gels (9,18). To examine the role of LRAP35a in DC migration, we also knocked down LRAP35a expression in Dock8 ϩ/Ϫ BM-derived DCs (Fig. 5A). Although knockdown of LRAP35a did not affect DC migration in a 2D setting, it resulted in reduced motility of Dock8 ϩ/Ϫ BM-derived DCs in collagen gels (Fig. 5B). Similar results were obtained by overexpressing LRAP35a N mutant in Dock8 ϩ/Ϫ BM-derived DCs (Fig. 5C). These results indicate that the  DOCK8-LRAP35a signaling axis also acts in DCs to promote migration within 3D environments.
Identification of the DOCK8 Region Critical for LRAP35a Association-To determine the region of DOCK8 critical for LRAP35a binding, we expressed WT or mutant DOCK8 in HEK-293T cells with LRAP35a and analyzed their interaction. Although WT DOCK8 and its mutant lacking the N-terminal 730 amino acid residues (DOCK8-⌬N⌬DHR1) comparably bound to LRAP35a, such interaction was hardly detected when the DOCK8 DHR2 domain was deleted (DOCK8-⌬DHR2, Fig. 6A). Consistent with this, an association between DOCK8 and LRAP35a was observed when the DOCK8 DHR2 domain alone was co-expressed with LRAP35a (Fig. 6B), suggesting that DOCK8 associates with LRAP35a through the DHR2 domain. This association seemed to be direct, as recombinant protein encoding the His-SUMO-tagged DOCK8 DHR2 domain exhibited definite binding to GST fusion LRAP35a but not the LRAP35a N mutant in vitro (Fig. 6C).
The DHR2 domain is composed of three lobes: lobe A, lobe B, and lobe C (21). Of these, lobe B of the DHR2 domain was found to be important for LRAP35a association (Fig. 6, D and E). Lobes B and C of DOCK8 are known to mediate Cdc42 binding (9). To examine whether Cdc42 binding affects the interaction between DOCK8 and LRAP35a, we expressed full-length DOCK8 with a VA mutation in the DHR2 domain. Although this mutant failed to bind to Cdc42, the VA mutation did not affect the interaction of DOCK8 with LRAP35a (Fig. 6F). In addition, pulldown experiments using GST fusion Cdc42 showed that DOCK8, LRAP35a, and Cdc42 were able to form a trimolecular complex (Fig. 6G). These results indicate that DOCK8 associates with LRAP35a through lobe B of the DHR2 domain, independent of Cdc42 binding. However, no association with LRAP35a was found when lobe B of the DOCK7 DHR2 domain was analyzed (Fig. 6E).

Discussion
Accumulating evidence indicates that DOCK8 regulates the migration and activation of various subsets of leukocytes (9 -18). However, how DOCK8 regulates these processes has been unclear. In this study, we showed that, in the absence of DOCK8, macrophages exhibit a migration defect even in a 2D setting. Although DOCK8 deficiency did not affect global Cdc42 activation, rescue experiments revealed that the Cdc42 GEF activity of DOCK8 is required for macrophage migration. We found that DOCK8 regulates the phosphorylation of MLC2 in chemotaxing macrophages. Mechanistically, DOCK8 associates with LRAP35a and facilitates MLC2 phosphorylation by MRCK␣. Interestingly, the association with LRAP35a was not found with other Cdc42 GEFs, such as ␤PIX and DOCK7 (28,29). Therefore, it is likely that the interaction with LRAP35a makes DOCK8 a special GEF that links Cdc42 activation to actomyosin dynamics during macrophage migration (Fig. 7).
The role of the DOCK8-LRAP35a signaling axis in other leukocyte subsets is also important. In this study, we showed that neither DOCK8 nor LRAP35a deficiency affected the DC migration on 2D surfaces. Although the precise reason is currently unknown, this discrepancy is likely to result from differences in the adhesiveness between macrophages and DCs.   However, we found that the migration speed of Dock8 ϩ/Ϫ DCs in collagen gels was also reduced to the level of Dock8 Ϫ/Ϫ DCs when LRAP35a expression was knocked down. Recent evidence indicates that NMII-mediated contractility is required for leukocytes to squeeze their nucleus through narrow gaps such as endothelial cell-cell junctions during extravasation and dense extracellular matrix scaffold during interstitial migration (30,31). Therefore, our results suggest that the DOCK8-LRAP35a axis also regulates interstitial DC migration by controlling actomyosin dynamics.
In conclusion, we showed here that DOCK8 regulates leukocyte migration by linking Cdc42 activation to actomyosin dynamics through LRAP35a interaction. Our findings provide novel insights into the mechanism of how DOCK8 regulates leukocyte migration.

Experimental Procedures
Mice-Dock8 Ϫ/Ϫ mice have been described previously (9). Mice were maintained under specific pathogen-free conditions in the animal facility of Kyushu University. All experiments were performed in accordance with the guidelines of the Committee of Ethics of Animal Experiments, Kyushu University.
Cell Preparation-For preparation of BM-derived macrophages, BM cells were cultured for 7 days in complete RPMI medium containing 10% heat-inactivated FCS (Nichirei Bioscience), 50 M 2-mercaptoethanol (Nacalai Tesque), 2 mM L-glutamine (Life Technologies), 100 units/ml penicillin (Life Technologies), 100 g/ml streptomycin (Life Technologies), 1 mM sodium pyruvate (Life Technologies), and MEM non-essential amino acids (Life Technologies) plus M-CSF (20 ng/ml, Pepro-Tech). To generate BM-derived DCs, BM cells were cultured for 7 days in complete RPMI medium supplemented with GM-CSF (10 ng/ml, PeproTech) and purified with anti-CD11c antibody-conjugated microbeads (Miltenyi Biotec). To isolate alveolar macrophages, mice were challenged with intratracheal instillations of Escherichia coli LPS (20 g/mouse). Mice were then killed 96 h later, and the bronchoalveolar lavage fluid was collected as described previously (32). For preparation of TGelicited peritoneal macrophages, Dock8 ϩ/Ϫ or Dock8 Ϫ/Ϫ mice were injected intraperitoneally with 200 l of sterile 3% TG broth solution (Nissui) to elicit an inflammatory response and recruit the peritoneal macrophages. Mice were sacrificed 96 h later by cervical dislocation, and peritoneal exudate cells were harvested and maintained in complete RPMI medium at 37°C. Three hours later, the adherent cells were collected and used as peritoneal macrophages.
Chemotaxis Assay-BM-derived macrophages or peritoneal macrophages were allowed to migrate under a CCL2 (10 g/ml, Wako) gradient in an EZ-TAXIScan chamber (Effector Cell Institute). Phase-contrast images of chemotaxing cells were acquired at 1-min intervals over 60 min. To assess the migration of BM-derived DCs, phase-contrast images of cells migrating toward CCL21 (250 g/ml, R&D Systems) were acquired at 30-s intervals over 30 min. EZ-TAXIScan chemotaxis assays were performed as described previously (9,33). TAXIScan-FL (Effector Cell Institute) was used when cells were transfected with a plasmid encoding the GFP fusion protein or Block-iT fluorescent oligo (Invitrogen). Images were imported as stacks to the ImageJ software program (National Institutes of Health). The velocity, directionality, and y-forward migration index were analyzed using manual tracking and the chemotaxis and migration tools. 3D collagen gel chemotaxis assays were performed with -Slide Chemotaxis 3D (Ibidi) as described previously (9). Briefly, cells were mixed with 1.65 mg/ml of collagen type I (Corning) in RPMI medium supplemented with 4% FCS and cast in the chamber before assays. After polymerization of the lattice, images of cells migrating toward CCL21 (10 g/ml) were taken in a heated chamber every 2 min using an IX-81 inverted microscope (Olympus) equipped with a cooled charge-coupled device camera (CoolSNAP HQ, Roper Scientific), an IX2-ZDC laser-based autofocusing system (Olympus), and an MD-XY60100T-Meta automatically programmable XY stage (Sigma KOKI). Images were taken, and velocities were calculated with the manual tracking feature in the MetaMorph software program.
Immunofluorescence Microscopy-BM-derived macrophages migrating in an EZ-TAXIScan chamber were fixed with 4% paraformaldehyde and permeabilized with 0.2% Triton X-100 in PBS. Cells were then stained with Alexa Fluor 546-conjugated phalloidin (1:500, Invitrogen), DAPI (1:5000, Dojindo), and anti-NMIIA antibody (1:400, PRB-440P, Covance). To stain activated myosin, cells were fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and incubated with the phosphorylation-specific antibody against Ser-19 of MLC2 (1:400, 3671, Cell Signaling Technology) and Alexa Fluor 488conjugated phalloidin (1:500, Invitrogen) for 1 h. After being washed with PBS, sections were incubated with appropriately FIGURE 7. Schematic of the role of DOCK8 in leukocyte migration. In this model, DOCK8 functions not only as a Cdc42 GEF but also as a scaffold protein that links Cdc42 activation to actomyosin dynamics through its interaction with LRAP35a. KIM, kinase inhibitory motif; CRIB, Cdc42/Rac-interactive binding site. labeled secondary antibodies. A microscopic analysis was performed with a laser-scanning confocal microscope (LSM 510 Meta, Carl Zeiss) with Plan-Apochromat ϫ63/1.4 oil objective at room temperature. The acquisition software program AIM (Carl Zeiss) was used. In some experiments, fluorescence intensity was measured from the cell edge to the cell center using the ImageJ software program. The averages from multiple cells were calculated and plotted against distance from the leading edge.
Plasmids and Transfection-The genes encoding DOCK8, DOCK7, LRAP35a, MRCK␣, and ␤PIX and their mutants with appropriate tags were created by PCR and subcloned into the pcDNA (Invitrogen) or pCI (Promega) vector. These plasmids were transfected into HEK-293T cells with polyethylenimine. The pMX vector was used to generate the plasmid encoding GFP or GFP-tagged MLC2. The pMX-MLC2-GFP or pMXs-GFP vector was transfected into platinum-E packaging cells with FuGENE 6 transfection reagent (Promega). 48 h after transfection, the cell culture supernatants were harvested and supplemented with Polybrene (5 g/ml) and M-CSF (20 ng/ml) to infect macrophages. The pCI-GFP vector encoding DOCK8, LRAP35a, and/or their mutants was transfected into macrophages or DCs by using Amaxa Nucleofector TM II (Lonza) in accordance with the instructions of the manufacturer. The genes encoding the DOCK8 DHR2 domain (residues 1633-2071) and its mutant were subcloned into the pET-SUMO vector to express fusion proteins with the His tag at their N termini (9). The GST fusion recombinant Cdc42, LRAP35a, and its mutants were expressed with the pGEX6P-1 vector (GE Healthcare).
RT-PCR-The cells were homogenized in Isogen (Nippon Gene). The RNA was extracted with chloroform, precipitated with isopropanol, and washed with 75% ethanol. The resulting RNA was resolved in diethyl pyrocarbonate-treated water. RNA samples were treated with RNase-free DNase I (Invitrogen), reverse-transcribed with oligo(dT) and SuperScript III reverse transcriptase (Invitrogen), and then amplified by PCR with specific primers.
Knockdown of LRAP35a and MRCK␣-The siRNA Mm_ 1520402A15Rik_1 (Qiagen, SI00791763) was transfected into BM-derived macrophages or DCs to knock down LRAP35a expression in accordance with the instructions of the manufacturer. The irrelevant oligonucleotide SI03650318 was used as a control. 48 h after transfection, cells were harvested for a chemotaxis assay and/or immunoblotting. The knockdown efficiency was checked by RT-PCR. The following primers were used: LRAP35a, 5Ј-ATGGAGGGGACCGCGGAGT-3Ј and 5Ј-GCCCTGTCCTCAGTGGGGATA-3Ј; actin, 5Ј-TGGA-ATCCTGTGGCATCCATGAAAC-3Ј and 5Ј-TAAAACGC-AGCTCAGTAACAGTCCG-3Ј. For knockdown of MRCK␣ expression, the siRNA Mm_Cdc42bpa_3 (Qiagen, SI02771769) was transfected into BM-derived macrophages in accordance with the instructions of the manufacturer. The irrelevant oligonucleotide SI03650318 was used as a control. 36 h after transfection, cells were harvested for immunoblotting. The knockdown efficiency was checked by immunoblotting using anti-MRCK␣ antibody (sc-374568, 1:1000, Santa Cruz Biotechnology).
In Vitro GEF Assay-The genes encoding the DOCK8 DHR2 domain and its mutant were expressed in E. coli Arctic Express (Stratagene) as fusion proteins with the His-SUMO tag at their N termini for purification with nickel-nitrilotriacetic acid-agarose chromatography (Qiagen). The GST fusion Cdc42 and GST alone were expressed in BL21 (Agilent Technologies) and purified on glutathione-Sepharose TM 4B (GE Healthcare) immediately before use. The assays consisted of GST fusion Cdc42 (10 M), BODIPY-FL GTP (2.4 M, Invitrogen), and the His-SUMO tagged recombinant WT DOCK8 DHR2 domain (residues 1633-2071) or its VA mutant (specified concentrations) in reaction buffer (20 mM MES-NaOH, 150 mM NaCl, 10 mM MgCl 2 , and 20 M GDP (pH 7.0)). For this purpose, GSTfusion Cdc42 (15.56 M) was loaded with GDP by incubation in reaction buffer in a total volume of 96.4 l on ice for 30 min and then mixed with BODIPY-FL GTP (0.1 mM, 3.6 l) and allowed to equilibrate at 30°C for 3 min. Recombinant DOCK8 DHR2 protein was equilibrated in reaction buffer in a total volume of 50 l for 30 min at room temperature. The reaction was initiated by mixing GDP-loaded Cdc42/BODIPY-FL GTP (100 l) and the recombinant DOCK8 DHR2 domain (50 l) in a final volume of 150 l and incubating at 30°C. The change in the BODIPY-FL-GTP fluorescence (excitation, 488 nm, emission, 514 nm) was monitored for 20 min using an EnSpire multimode plate reader (PerkinElmer Life Sciences). Data were fitted using the curve fitting function in the GraphPad Prism 5 software program (GraphPad Software), and the initial slope during the first 10 s (in relative fluorescent unit [RFU] per second) was calculated and used for comparison of the GEF activity.
Cdc42 Activation Assays-BM-derived macrophages suspended in Hanks' balanced salt solution were stimulated with CCL2 (100 ng/ml). Aliquots of the cell extracts were kept for total lysate controls, and the remaining extracts were incubated with the GST fusion Cdc42-binding domain of PAK1 (Merck Millipore) at 4°C for 60 min. The bound proteins and the total lysate control were analyzed by SDS-PAGE, and the blots were probed with anti-Cdc42 antibody (1:1000, 610928, BD Biosciences).
Statistical Analyses-Statistical analyses were performed using the GraphPad Prism software program. We first calculated the Gaussian distribution of the data using the Kolmogorov-Smirnov test. When two groups were compared, two-tailed Student's t test (Gaussian distribution) or Mann-Whitney test (no Gaussian distribution) was used. p Ͻ 0.05 was considered significant.
Author Contributions-A. S., F. S., and D. S. performed functional and immunohistochemical analyses and analyzed the data. T. U. and M. U. performed biochemical analyses. A. S., T. U., and T. H. helped write the manuscript. Y. F. conceived the project, interpreted the data, and wrote the manuscript.