The Gα13-Rho Signaling Axis Is Required for SDF-1-induced Migration through CXCR4*

The CXC chemokine stromal cell-derived factor-1α (SDF-1) binds to CXCR4, a seven-transmembrane G protein-coupled receptor that plays a critical role in many physiological processes that involve cell migration and cell fate decisions, ranging from stem cell homing, angiogenesis, and neuronal development to immune cell trafficking. CXCR4 is also implicated in various pathological conditions, including metastatic spread and human immunodeficiency virus infection. Although SDF-1-induced cell migration in CXCR4-expressing cells is sensitive to pertussis toxin treatment, hence involving heterotrimeric G proteins of the Gi family, whether other G proteins participate in the chemotactic response to SDF-1 is still unknown. In this study, we took advantage of the potent chemotactic activity of SDF-1 in Jurkat T-cells to examine the nature of the heterotrimeric G protein subunits contributing to CXCR4-mediated cell migration. We observed that whereas Gi and Gβγ subunits are involved in SDF-1-induced Rac activation and cell migration, CXCR4 can also stimulate Rho potently leading to the phosphorylation of myosin light chain through the Rho effector, Rho kinase, but independently of Gi. Furthermore, we found that Gα13 mediates the activation of Rho by CXCR4 and that the functional activity of both Gα13 and Rho is required for directional cell migration in response to SDF-1. Collectively, our data indicate that signaling by CXCR4 to Rho through Gα13 contributes to cell migration when stimulated by SDF-1, thus identifying the Gα13-Rho signaling axis as a potential pharmacological target in many human diseases that involve the aberrant function of CXCR4.

other heterotrimeric G proteins are also required for cell migration and, if so, the nature of their downstream targets is not fully understood. Here, we took advantage of the fact that Jurkat T-cells express CXCR4 receptors endogenously, and thus migrate readily in response to SDF-1, to begin dissecting the signaling network by which CXCR4 controls directional cell motility. We found that whereas G i regulates cell migration by activating Rac through G␤␥ heterotrimeric G protein subunits, CXCR4 utilizes the ␣ subunit of G 13 to activate Rho and that the functional activity of G i and G 13 , hence the coordinated activation of Rac and Rho, are both required for CXCR4-induced cell migration in response to SDF-1.
Cell Lines and Transfection-Jurkat cells were cultured in RPMI 1640 (Sigma) supplemented with 10% fetal calf serum (BioWhittaker, Walkersville, MD). Human epithelial kidney 293-T cells were grown and maintained in Dulbecco's modified Eagle's medium (Sigma) containing 10% fetal calf serum. Cells were transfected by Lipofectamine-PLUS (Invitrogen) following the manufacturer's directions.
Lentivirus Infection-Lentiviral stocks were prepared and titrated as previously reported using 293T cells as the packaging cells (25). Jurkat cells were incubated with viral supernatants for 12 h. After that, the cells were washed twice with phosphatebuffered saline and returned to normal growth medium.
Flow Cytometric Analysis-Cells were harvested and washed three times with phosphate-buffered saline. Following incubation with biotin-conjugated anti-CXCR-4 antibody or isotype control antibody for 60 min at room temperature, the cells were treated with streptavidin-phycoerythrin-conjugated IgG (Vector Laboratories, Burlingame, CA) for 30 min at room temperature and analyzed with a BD Biosciences flow cytometer.
Chemotaxis Assay-Chemotaxis assay was determined with a 48-well Boyden chamber (NeuroProbe, Gaithersburg, MD) using a polyvinyl pyrrolidone-free polycarbonate filter with a 5-m pore size (Nuclepore; Corning Costar, Acton, MA). Filters were not coated with extracellular matrix molecules to prevent the influence of cell adhesion to the chemotactic response to SDF-1. Briefly, 50 l of Jurkat cells transfected by Renilla luciferase with or without other constructs were added to the upper chamber, and a similar volume of cell suspension was kept as transfection efficiency control. The chemoattractant was added to the lower chamber. Cells migrated into the lower chamber were collected after 4 h of incubation, and Renilla luciferase activities present in cellular lysates were assayed using the dual luciferase reporter system. In each case, migration was calculated as the ratio between the luciferase activity in migrating cells (lower chamber) and the total luciferase activity in the 50-l cell suspension.
Rho GTPase Pulldown Assay-In vivo Rho and Rac activity was assessed by a modified method described elsewhere (21). Briefly, after serum starvation overnight, cells were treated as indicated and lysed at 4°C in a buffer containing 20 mM Hepes, pH 7.4, 0.1 M NaCl, 1% Triton X-100, 10 mM EGTA, 40 mM ␤-glycerophosphate, 20 mM MgCl 2 , 1 mM Na 3 VO 4 , 1 mM dithiothreitol, 10 g/ml aprotinin, 10 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride. Lysates were incubated with glutathione S-transferase-rhotekin-Rho binding domain previously bound to glutathione-Sepharose beads for Rho activity assay or with a purified, bacterially expressed glutathione S-transferase fusion protein containing the cdc42/Rac-interacting binding (CRIB) domain of PAK1 previously bound to glutathione-Sepharose beads for Rac activity. Associated GTPbound forms of Rho or Rac were released with protein loading buffer and analyzed by Western blot analysis using a monoclonal antibody against Rho or Rac.
Immunoblot Analysis-Cells were lysed in lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% Nonidet P-40) supplemented with protease inhibitors (0.5 mM phenylmethylsulfonyl fluoride, 1 l/ml aprotinin and leupeptin) for 15 min at 4°C. Equal amounts of protein were subjected to SDS-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride membrane (Immobilon P; Millipore). The membranes were then incubated with the appropriate antibodies.

SDF-1 Induces Jurkat Cell Migration through CXCR4-To
investigate the mechanism underlying the ability of SDF-1 to promote cell migration, we took advantage of the Jurkat T cell line that expresses CXCR4 endogenously (26). We first confirmed CXCR4 expression on Jurkat cells by fluorescence-activated cell sorter analysis. Saos-2 cells, which do not express of CXCR4, were used as a negative control. As shown in Fig. 1A, we observed remarkable expression of CXCR4 on Jurkat cells, whereas no expression of CXCR4 was detected on Saos-2 cells (not shown). We then determined the functional activity of CXCR4 in Jurkat cells by analyzing the activation of ERK, a typical response elicited by GPCRs (20) by SDF-1. Exposure of Jurkat cells to SDF-1 resulted in the fast accumulation of the phosphorylated active form of ERK, which peaked at ϳ5 min after stimulation and decreased thereafter. Pretreatment of cells with AMD3100, a specific CXCR4 inhibitor (27), prevented the phosphorylation of ERK when provoked by SDF-1 (Fig. 1B), thus together demonstrating that Jurkat cells express functional CXCR4 receptors. We next evaluated whether SDF-1 could induce the migration of Jurkat cells using a modified Boyden Chamber system in which cell migration can be easily quantified using a luciferase-based assay. Using this system, we observed that Jurkat cells migrate efficiently toward wells containing various concentrations of SDF-1, reaching a maximum already at a concentration of 10 ng/ml of this chemoattractant (Fig. 1C). The treatment with AMD3100 prevented the SDF-1-induced cell migration in a dose-dependent manner (Fig. 1D). These data indicated that SDF-1 can induce a remarkable migratory response in Jurkat cells through CXCR4.
G i and G␤␥ Are Involved in the Migration of Jurkat Cells Induced by SDF-1-PTX, which ADP ribosylates G␣ i , thereby uncoupling it from receptor activation (28), was used to investigate the role of G i in the process of cell migration stimulated by SDF-1. The effect of PTX on the phosphorylation of ERK was used as positive control. Treatment of Jurkat cells with PTX resulted in a decrease of phosphorylation of ERK and cell migration in a dose-dependent fashion (Fig. 2, A and B). This confirmed that G i participates in the migratory response initiated by CXCR4. As receptor stimulation of G i induces the dissociation of G␤␥ subunits from GTP-bound G i (29), we next evaluated the contribution of G␤␥ to SDF-1-induced cell migration by the overexpression of G␣ transducin, which blocks G␤␥ function by quenching these G protein subunits upon their release from G i (20). Indeed, expression of G␣ transducin in Jurkat cells (Fig. 2C) prevented the phosphorylation of ERK induced by SDF-1 (Fig. 2D). In contrast, the phosphoryla- Jurkat cells were serum starved, treated (ϩ) or not (Ϫ) with AMD3100, and stimulated with SDF-1 (10 ng/ml) for the indicated time. The total (ERK) and phosphorylated forms (p-ERK) of ERK1 and ERK2 were detected in cellular lysates by Western blot analysis with the appropriate antibodies. C, SDF-1 induces the migration of Jurkat cells. Serum-free growth medium containing various concentrations of SDF-1 was added into the lower wells of Boyden chambers, and Jurkat cells transfected with Renilla luciferase were loaded into the upper wells. After 4 h, the cells that migrated into the lower wells were collected. Migration was measured as the ratio of the Renilla luciferase activity in migrating cells versus the total Renilla luciferase activity in the cells added in the upper chamber of each well and expressed as -fold increase with respect to cell migration toward serum-free growth medium, which was taken as 1. Values are the average Ϯ S.E. of triplicate samples from a typical experiment. Similar results were obtained in three additional experiments. D, AMD3100 inhibits the migration induced by SDF-1. Jurkat cells were treated with the indicated concentrations of AMD3100 and their migration toward wells containing SDF-1 (10 ng/ml) measured and represented as for panel C.

FIGURE 2. PTX-sensitive G proteins of the G i family are required for SDF-1-induced migration of Jurkat T-cells. A role for G␤␥ subunits.
A, PTX inhibits the phosphorylation of ERK stimulated by SDF-1. Jurkat cells pretreated overnight with serum-free medium with or without different concentrations of PTX were stimulated with SDF-1 (10 ng/ml) for 5 min. Cells were lysed and used for Western blot analysis of phosphorylated and total ERK. B, PTX blocks the migration of Jurkat cells evoked by SDF-1. After incubating with different concentrations of PTX overnight, Jurkat cells were suspended in serum-free growth medium and added into the upper wells of the Boyden chambers. Serum-free growth medium with or without SDF-1 (10 ng/ml) was loaded into the lower wells as chemoattractant. Data are represented as in Fig. 1C. C, expression of transducin in Jurkat cells. Control and transducin (G␣ t )-transfected Jurkat cells were analyzed by Western blotting using antibodies to transducin or ␣-tubulin as a control. D, expression of transducin decreases the phosphorylation of ERK induced by SDF-1. Jurkat cells were transfected with transducin, serum starved overnight, and lysed after stimulating with SDF-1 (10 ng/ml) or PMA (50 nM) for 5 min. Western blot analysis of ERK phosphorylation was performed as above. E, transducin inhibits the migration of Jurkat cells induced by SDF-1. Jurkat cells transfected with different concentrations of transducin were subjected to migration assays using SDF-1 as the chemoattractant as described above. tion of ERK elicited by PMA through protein kinase C (30) remained unaffected and served as a negative control (Fig. 2D). G␣ transducin expression resulted in a dramatic inhibition of the migratory response toward SDF-1 (Fig. 2E), thus indicating that G␤␥ subunits, when released from G i , are involved in SDF-1-induced cell migration through CXCR4.
Signaling through G i Promotes the Migration of Jurkat Cells by Stimulating Rac-Because the small GTP-binding protein Rac plays a central role in the regulation of the actin-based cytoskeleton and cell movement (31), we then investigated whether this Rho-related GTPase is involved in the G i -mediated migration of Jurkat cells. As shown in Fig. 3A, by a pulldown assay we observed that stimulation of Jurkat cells with SDF-1 provoked the rapid accumulation of GTP-bound Rac, which reached a maximum at ϳ5 min. In parallel experiments, however, we did not detect the accumulation of the GTPbound form of Cdc42, a Rac-related GTPase (not shown). Pretreatment of Jurkat cells with PTX attenuated the Rac activity induced by SDF-1 (Fig. 3B). These data indicate that SDF-1 induces Rac activation through PTX-sensitive G proteins of the G i class. We next set out to explore whether Rac is necessary for the migratory response to SDF-1 by expressing in Jurkat cells the N-terminal region of PAK (p21-activated kinase), PAK-N, which includes its Cdc42/Rac-interactive binding motif, and PAK-NL2, which lacks two residues required for Rac binding and therefore serves as a specificity control (21). Fig. 3C shows the expression of the AU1-tagged forms of both PAK-N and PAK-NL2 after transfection of their expression vectors into Jurkat cells. Interestingly, as shown in Fig. 3D, expression of PAK-N dramatically reduced the migration of Jurkat cells induced by SDF-1, whereas PAK-NL2 did not exert any demonstrable effect (Fig. 3D). Collectively, our results suggest that G i participates in the SDF-1-induced migration of Jurkat cells by stimulating the activity of Rac, likely through G␤␥ subunits.
Rho Is Involved in Cell Migration Induced by SDF-1 through CXCR4-As cell migration often requires the coordinated activation of the small GTPases Rac and Rho, we next sought to explore whether Rho plays a role in the chemotactic response to SDF-1. We first examined whether SDF-1 promotes the activation of Rho. As shown in Fig. 4A, SDF-1 stimulation provoked a profound activation of Rho, as judged by the use of pulldown assays. To address whether Rho is involved in cell migration stimulated by SDF-1, we used both treatment with C3 toxin, by which ADP ribosylates and inhibits Rho (32), and the expression of an AU5-tagged negative dominant mutant of Rho, Rho N19. Treatment with C3 toxin from 0.1-5 ng/ml inhibited the migration of Jurkat cells toward wells containing SDF-1 (Fig.  4B). Similarly, the expression of the AU5-tagged form of Rho N19 in Jurkat cells (Fig. 4C) dramatically reduced the migration of these cells when induced by SDF-1 (Fig. 4D), while in control experiments the dominant negative mutant of Rac, but not of Cdc42, also prevented Jurkat cell migration (not shown). These data demonstrate that SDF-1 can activate Rho and that in turn Rho is involved in cell migration induced by this potent chemokine. To further investigate whether Rho activation in response to SDF-1 is indeed mediated through CXCR4, we limited the expression of this GPCR by RNA-interfering approaches (Fig. 4E). Our data showed that the knock down of the expression of CXCR4 in Jurkat cells resulted in a reduction in the ability of SDF-1 to promote both Rho activation and migration (Fig. 4, F and G).
Rho stimulation leads to the activation of Rho kinase, which promotes the accumulation of phospho-MLC by phosphorylating MLC directly or by inhibiting MLC phosphatase, thereby contributing to cell migration by regulating acto-myosin contraction (33). To explore whether Rho kinase is an important downstream effector of Rho activation in response to SDF-1, we treated Jurkat cells with the Rho kinase inhibitor Y27632 (34) and observed that SDF-1 stimulates the phosphorylation of MLC potently (Fig. 4H), which was abolished by the treatment with Y27632. Furthermore, Y27632 prevented the migration of Jurkat cells toward SDF-1 (Fig. 4I). These findings suggest that the stimulation of Rho by SDF-1 through CXCR4 results in Rho kinase activation and consequent MLC phosphorylation and that these downstream events initiated by Rho participate in cell migration toward SDF-1.
G␣ 12/13 Is Involved in SDF-1-induced Cell Migration, and This Effect Is Mediated by Rho-As both G i and Rho activity were required for SDF-1-induced cell migration, we asked whether Rho functions downstream of G i . However, as shown in Fig. 5A, treatment of Jurkat cells with PTX failed to inhibit Rho activation in response to SDF-1 and instead caused a slight enhancement of Rho activity (Fig. 5A). As G␣ 12 and G␣ 13 can promote the activation of Rho elicited by GPCRs through directly binding to the RGS domain of RGS-containing Rho guanine nucleotide exchange factors (GEFs) (22,35), we then examined whether the G␣ 12/13 -Rho signaling axis is required for cell migration when induced by SDF-1. We transfected Jurkat cells with a chimeric molecule encoding GFP fused to the RGS domain of PDZ-RhoGEF, which binds to the activated forms of G␣ 12/13 and thus behaves as a dominant negative mutant for G␣ 12/13 (25). Expression of this GFP-RGS chimera in Jurkat cells (Fig. 5B) inhibited the migration elicited by SDF-1 (Fig. 5D), suggesting that G␣ 12/13 activation is required for the chemotactic response to SDF-1. We then set out to explore whether the effect of G␣ 12/13 on the migration of Jurkat cells is mediated by Rho. As shown in Fig. 5C, Jurkat cells transfected by various concentrations of GFP-RGS exhibited reduced Rho activity after SDF-1 stimulation as compared with control-transfected cells. These data provided evidence that G␣ 12/13 is involved in cell migration when induced by SDF-1 and that this effect is likely mediated by Rho.
As 293T cells do not express CXCR4 (not shown), we then set out to reconstitute the CXCR4-G␣ 13 signaling axis by the

. G␣ 12/13 is involved in the migration of Jurkat cells induced by SDF-1. A role for G␣ 12/13 in Rho activation.
A, inhibition of the activity of G i by PTX fails to prevent Rho activation. Jurkat cells treated with different concentrations of PTX overnight were stimulated with SDF-1 (10 ng/ml; 5 min), and Rho activity in cell lysates was analyzed as described above. B, expression of the GFP-tagged RGS domain of PDZ-RhoGEF in Jurkat cells. Western blot analysis with antibodies to GFP or ␣-tubulin was performed in lysates from Jurkat cells transfected with the GFP-tagged RGS domain of PDZ-RhoGEF (GFP-RGS) or vector control. C, inhibition of G␣ 12/13 suppresses the activation of Rho by SDF-1. Jurkat cells transfected with GFP-RGS were stimulated with SDF-1 (10 ng/ml) for 5 min and Rho activity measured in cell lysates by pulldown analysis. D, inhibition of G␣ 12/13 diminishes SDF-1-induced cell migration. Jurkat cells transfected with different amounts of GFP-RGS were used for migration assays toward SDF-1 (10 ng/ml) as above. E and F, reconstitution of the CXCR4-G␣ 13 -signaling axis in 293T cells. E, time course of ERK phosphorylation in response to SDF-1 in 293T cells transfected with expression vectors for GFP or CXCR4. Cells treated with 10 ng/ml SDF-1 were collected at the indicated time for immunoblot analysis. F, G␣ 13 is involved in Rho activation in response to SDF-1 in CXCR4-transfected cells. 293T cells transfected with GFP or CXCR4 with or without GFP RGS were lysed after treatment with SDF-1 and used for pulldown assay. ectopic expression of this GPCR upon transfection of its expression plasmid in 293T cells. As a control, although SDF-1 stimulation of 293T cells transfected with a vector control (GFP) did not cause ERK activation, the phosphorylated form of ERK was readily detected in 293T cells transfected with CXCR4 expression plasmids after SDF-1 treatment (Fig. 5E). Similarly, activation of CXCR4-transfected 293T cells led to a clear increase in the level of GTP-bound Rho, which was prevented by the co-expression of the RGS domain of PDZ-RhoGEF (Fig.  5F), hence supporting the emerging notion that G␣ 12/13 is involved in Rho activation downstream from CXCR4.
Knock Down of G␣ 13 in Jurkat Cells Inhibits Cell Migration and Rho Activation Induced by SDF-1-To address the direct role of G␣ 12/13 in the migration of Jurkat cells mediated by CXCR4, we set out to knock down G␣ 12/13 by using RNA interference approaches. We first examined the expression of G␣ 12 and G␣ 13 in Jurkat cells using 293T cells transfected with G␣ 12 or G␣ 13 , respectively, as positive controls. As shown in Fig. 6A, Jurkat cells only express G␣ 13 , while expression of G␣ 12 is undetectable. We then designed a shRNA sequence targeting G␣ 13 and expressed it stably in Jurkat cells by a recently developed lentiviral system (24). As shown in Fig. 6B, G␣ 13 expression was decreased 6 days after infection of Jurkat cells with a lentivirus carrying G␣ 13 shRNA and was almost eliminated 10 days after infection, whereas a control GFP-expressing lentivirus had no demonstrable effect. As shown in Fig. 6C, the activation of Rho in Jurkat cells mediated by CXCR4 was significantly blocked after G␣ 13 knock down. In addition, phosphorylation of ERK elicited by SDF-1 was slightly decreased, suggesting an unexpected role for G 13 in ERK activation by SDF-1. Remarkably, we also found that knock down of the expression of G␣ 13 reduced the migration mediated by CXCR4, in alignment with the result obtained by the G␣ 12/13 dominant negative mutant (Fig. 6D). Taken together, our data suggest that CXCR4 is coupled to G 13 , in addition to G i , and that activation of Rho through the ␣ subunit of G 13 , G␣ 13 , is required to induce cell migration in response to SDF-1.

DISCUSSION
In the present study, we used Jurkat T-cells that express CXCR4 endogenously as an experimental model system to examine the contribution of heterotrimeric G proteins and their downstream targets in cell migration induced by SDF-1. Using a simple luciferase-based system to monitor cell migration, we first observed that SDF-1 elicits a robust chemotactic response through CXCR4 in Jurkat cells. This response was abolished by treatment with PTX, confirming the key role for G proteins of the G i family in lymphocyte migration toward chemokine gradients (36). In this regard, we observed that G␤␥ subunits, when released from G i , are strictly required for CXCR4-mediated cell migration and that the chemotactic response to SDF-1 involves the coordinated activation of the small GTPases Rac and Rho by CXCR4. Of interest, whereas the activation of Rac was mediated by G i , likely though its G␤␥ subunits, the activation of Rho was insensitive to PTX treatment, thus suggesting that CXCR4 may utilize G proteins distinct from G i to stimulate Rho. Indeed, by a combination of dominant negative and RNA interference approaches, we now show that G proteins of the G␣ 12/13 family are also required to induce cell migration in response to SDF-1 and that G␣ 13 links CXCR4 to the activation of Rho by SDF-1 in Jurkat T cells.
Small GTP-binding proteins of the Rho family, including Rho, Rac, and Cdc42, play a central role in regulating the dynamic organization of the actin-based cytoskeleton in mammalian cells (37). During the process of cell migration, each member of this Rho GTPase family plays a distinct role. Rho is important for regulating the formation of contractile actin-myosin filaments, which form stress fibers, and for maintaining focal adhesions at the rear of the migrating cells, whereas Rac is involved in forming actin-rich membrane ruffles, referred to as lamellipodia, at the leading edge of the migrating cells and is recognized to be a driving force for cell migration (38). Cdc42 has been demonstrated to be critical in regulating cell polarity and filopodia formation, thereby controlling the direction of the cell migration (39). Hence, the coordinated activation of Rho GTPases represents a key regulatory event during the migration of cells toward a chemoattractant gradient. How this  13 . The expression of G␣ 12 and G␣ 13 in Jurkat cells was analyzed by Western blot with heterotrimeric G protein ␣ subunit-specific antibodies, using antibodies to ␣-tubulin as a loading control. 293T cells transfected with empty vector (Ϫ) or G␣ 12 and G␣ 13 expression vectors were used as negative and positive controls, respectively. B, knock down of G␣ 13 by lentiviruses carrying G␣ 13 shRNA. Jurkat cells infected with lentiviruses carrying GFP or G␣ 13 shRNA were lysed at the indicated times after infection and used for Western blot analysis of G␣ 13 expression. Western blot for ␣-tubulin was used as a loading control. C, knock down of the expression of G␣ 13 partially decreases the phosphorylation of ERK and nearly abolishes the activation of Rho induced by SDF-1. The accumulation of phosphorylated ERK and GTP-bound Rho in response to SDF-1 (10 ng/ml; 5 min) was examined as described above in Jurkat cells 10 days after infection with lentiviruses carrying GFP or G␣ 13 shRNA. D, knock down of G␣ 13 inhibits the migration of Jurkat cells stimulated by SDF-1. The migration of Jurkat cells to wells containing SDF-1 (10 ng/ml) was examined as described above 10 days after infection with lentiviruses carrying GFP or G␣ 13 shRNA and normalized by their respective migration to serum-free growth medium. DECEMBER 22, 2006 • VOLUME 281 • NUMBER 51 coordination is achieved is under current intense investigation (37,40,41).

CXCR4 Signals to Rho through G␣ 13
In Jurkat T cells, we observed that SDF-1 induces the rapid activation of both Rac and Rho, but not Cdc42, in line with prior studies using human peripheral blood lymphocytes (42,43). Of interest, PTX treatment suppressed the activation of Rac while slightly enhancing Rho activity. This observation is consistent with a role for G i in signaling from chemokine receptors to Rac (42) and prior studies that demonstrated that the activity of Rho can be negatively regulated by Rac (44,45). However, the nature of the molecular mechanism by which G i regulates Rac is at the present not fully understood. A recent study suggests that Vav1, a Rac GEF that is activated upon tyrosine phosphorylation (46), changes its localization and activity upon SDF-1 stimulation of lymphocytes (47). Furthermore, given the fact that CXCR4 activates phosphatidylinositol 3-kinase (18) and the Tec family tyrosine kinase ITK through Src (19), it is quite likely that CXCR4 may promote the activation of Rac through the sequential stimulation of Src, phosphatidylinositol 3-kinase, and ITK and the consequent tyrosine phosphorylation and phosphatidylinositol 3-kinase-dependent activation of Vav1. However, it is still possible that, upon SDF-1 binding to CXCR4, G␣ i or free G␤␥ may also promote the direct activation of G protein-coupled Rac GEFs such as Tiam1 (48) and P-REX2 (49), both of which are expressed in Jurkat cells (not shown). Thus, the relative contribution of each of these Rac GEFs to the activation of Rac by CXCR4 and lymphocyte migration warrants further investigation.
On the other hand, the migratory response of Jurkat T cells toward SDF-1 required the functional activity of Rho, but the activation of this GTPase by CXCR4 was not mediated by G i . This prompted us to explore whether additional G protein ␣ subunits may link CXCR4 to Rho. In this regard, G␣ 12 and G␣ 13 , members of the G␣ 12 family of G proteins, are known to stimulate signaling pathways involved in cell growth and neoplastic transformation (50,51), but accumulating evidence also supports a role for these G protein ␣ subunits in the regulation of cell migration. For example, concertina, an ortholog of G␣ 12/13 , is critically required for cell movement during Drosophila embryogenesis (52), and G␣ 13 knockout embryonic fibroblasts exhibit reduced migratory response to thrombin and lysophosphatidic acid (53). Hence, in the present work, we found that G␣ 12/13 plays a key role in the chemotactic response initiated by SDF-1 through CXCR4. Furthermore, the results obtained by reconstituting the CXCR4-G␣ 13 signaling axis in 293T cells and the knock down of G␣ 13 by the infection of Jurkat cells with lentiviruses carrying G␣ 13 shRNA revealed that CXCR4 stimulates Rho activation through G␣ 13 . Thus, the remarkable biological responses elicited by CXCR4 may require the coordinated activation of Rac-dependent pathways through G i and biochemical routes regulated by Rho through the activation of G 13 . As GTP-bound G␣ 13 can activate Rho through RGS-containing Rho GEFs, p115 RhoGEF, PDZ-Rho-GEF, or LARG (22,35), these findings may also provide a molecular framework to help define the precise GEF utilized by CXCR4 to stimulate Rho in Jurkat T-cells and in many other cell types in which SDF-1 and CXCR4 regulate normal and aberrant cell migration.
The finding that CXCR4 utilizes a G␣ 12/13 -Rho signaling axis to promote cell migration may have broad implications as the SDF-1/CXCR4 signaling system plays a critical role in many physiological processes, such as axon guidance, stem cell homing, tissue damage repair, and hematopoietic cell trafficking (8,13,14,17), in which the contribution of G␣ 12/13 and Rho can now be investigated. Similarly, although signaling through G i by CXCR4 appears not to be necessary for the infection of lymphocytes by T-cell tropic (R4) HIV (10), the ability of this chemokine receptor to couple to G 13 , thereby stimulating Rho-dependent pathways that affect cell motility and nuclear events, may now help explain the still unknown post-entry mechanism by which signaling through CXCR4 contributes to HIV replication and viral transmission (reviewed in Refs. 2, 54). Furthermore, SDF-1 is produced by stromal cells in the lymph nodes and acts as a potent chemoattractant for many cancer cells that express CXCR4, including those from the breast, prostate, and lung, thereby promoting lymph node invasion and the subsequent metastatic spread of these highly prevalent cancers (14,16,36). Thus, the finding that CXCR4 signals through G␣ 13 to Rho in response to SDF-1 may provide novel therapeutic targets for pharmacological intervention in many pathological conditions, ranging from increasing the efficiency on bone marrow transplantation to treating cancer metastasis and HIV infection.