CCR7/CCL21 Migration on Fibronectin Is Mediated by Phospholipase Cγ1 and ERK1/2 in Primary T Lymphocytes*

CCR7 binds to its cognate ligand, CCL21, to mediate the migration of circulating naive T lymphocytes to the lymph nodes. T lymphocytes can bind to fibronectin, a constituent of lymph nodes, via their β1 integrins, which is a primary mechanism of T lymphocyte migration; however, the signaling pathways involved are unclear. We report that rapid (within 2 min) and transient phosphorylation of ERK1/2 is required for T cell migration on fibronectin in response to CCL21. Conversely, prevention of ERK1/2 phosphorylation by inhibition of its kinase, MAPK/MEK, prevented T lymphocyte migration. Previous studies have suggested that phospholipase Cγ1 (PLCγ1) can mediate phosphorylation of ERK1/2, which is required for β1 integrin activation. Paradoxically, we found that inhibition of PLCγ1 phosphorylation by the general PLC inhibitor U73122 was associated with a delayed and reduced phosphorylation of ERK1/2 and reduced migration of T lymphocytes on fibronectin. To further characterize the relationship between ERK1/2 and PLCγ1, we reduced PLCγ1 levels by 85% using shRNA and observed a reduced phosphorylation of ERK1/2 and a significant loss of CCR7-mediated migration of T lymphocytes on fibronectin. In addition, we found that inhibition of ERK1/2 phosphorylation by U0126 resulted in a decreased phosphorylation of PLCγ1, suggesting a feedback loop between ERK1/2 and PLCγ1. Overall, these results suggest that the CCR7 signaling pathway leading to T lymphocyte migration on fibronectin is a β1 integrin-dependent pathway involving transient ERK1/2 phosphorylation, which is modulated by PLCγ1.

stream signaling events to regulate the activation state of adenylyl cyclase or phospholipase C (PLC) (1). CCR7 (C-C chemokine receptor 7), a G protein-coupled receptor, is expressed on naive T lymphocytes and is required for migration into and within lymph nodes. CCR7 binds to two ligands, CCL19 and CCL21. CCL21 is expressed in high endothelial venules, the entry route into lymph nodes (2). However, it is currently unknown how T lymphocytes signal through CCR7 to mediate migration via ␤1 integrins in response to CCL21.
Integrins, which are key mediators of lymphocyte adhesion and migration, are heterodimeric adhesion proteins consisting of an ␣ and a ␤ subunit, for which they are named. ␤1 integrins, which are expressed by T lymphocytes, bind to the vascular cell adhesion molecule, collagen, and fibronectin (3). Fibronectin is a major component of the lymph node and is highly expressed in the cortex in the presence of fibroblastic reticular cells, which express CCL21 (4,5). This network provides an environment that allows CCR7-expressing T lymphocytes to migrate throughout the lymph node.
PLC␤ is a downstream target of G␤␥ signaling and is important for T cell migration (6). PLC␥ binds to and activates ␤1 integrins (7). Furthermore, PLC␥1 is required for adhesion to fibronectin and is important for ␤1 integrin-mediated cell migration (8,9). PLC␥1 was recently shown to be phosphorylated in response to CCR7 activation in head and neck cancer cells and was necessary for the migration of squamous cell carcinoma cells (10).
GPCRs can activate the downstream effector MAPKs to regulate chemotactic migration (11,12). In HEK293 cells, CCR7 transient transfectants stimulated with CCL19 rapidly activated ERK1/2 (extracellular signal-related kinase 1/2), whereas CCL21 had a minimal effect on ERK1/2 phosphorylation (13). In primary B lymphocytes that express endogenous CCR7, CCL21 mediated sustained activation of ERK1/2 for 10 min (11). These contrasting studies illustrate the importance of determining the role of ERK1/2 in the migration of primary T lymphocytes to CCL21. Therefore, we examined the molecular mechanisms required for ␤1 integrin activation and ␤1 integrin-mediated migration. We found that in primary T lymphocytes, CCL21 promotes the activation of PLC␥1 and the transient activation of ERK1/2, which are required for ␤1 integrin activation in response to CCL21. Loss of activation of either PLC␥1 or transient phosphorylation of ERK1/2 prevented migration via ␤1 integrins to CCL21. These results indicate that CCL21 activation of CCR7 promotes the activation of ␤1 integrins, leading to migration following phosphorylation of PLC␥1 and ERK1/2.

EXPERIMENTAL PROCEDURES
Mice, Primary Human Lymphocytes, and Reagents-Primary human T lymphocytes were isolated from volunteer donors under an approved protocol in accordance with the policies and procedures of the human subjects protection program at the University of Kansas Medical Center. U0126 (Calbiochem), U73122 and U73343 (Biomol), pertussis toxin (List Biological Laboratories), human PLC␥1 shRNA (OriGene), and speciesspecific CCL19 and CCL21 (R&D Systems) were purchased. The 12G10 anti-human activated ␤1 integrin antibody was generously provided by Dr. Martin Humphries (University of Manchester, Manchester, UK), and the MAB1959 functionblocking antibody was purchased (Millipore).
T Lymphocyte Isolation-Human blood was collected from volunteer donors, and peripheral blood mononuclear cells were isolated using a Ficoll-Paque gradient (GE Healthcare). Whole blood was transferred to 50-ml conical tubes and diluted at a 1:1 ratio with PBS ϩ Ca 2ϩ /Mg 2ϩ (Cellgro). In a separate 50-ml conical tube, 15 ml of Ficoll-Paque was layered with 35 ml of whole blood/PBS ϩ Ca 2ϩ /Mg 2ϩ and centrifuged for 20 min at 100 ϫ g without the brake. The middle layer containing lymphocytes was isolated, washed, and pelleted. T lymphocytes were negatively selected using the EasySep kit (Stemcell Technologies) according to the manufacturer's protocol. Lymphocytes were maintained in RPMI 1640 medium (Invitrogen), 10% heat-inactivated FBS (Hyclone), and 2 nM L-glutamine (Invitrogen) in a humidified atmosphere at 37°C and 5% CO 2 for use within 3 days of isolation.
Chemotaxis Assays-All chemotaxis assays were carried out using a 48-well chemotaxis chamber (Neuroprobe). Lower wells were loaded with 0 or 10 nM to 2 M CCL21 in serum-free RPMI 1640 medium and were separated from the upper wells containing the lymphocytes using a 5-m pore nitrocellulose membrane (Neuroprobe) that had been preincubated in either serum-free RPMI 1640 medium or 10 g/ml fibronectin (Sigma). Primary human T lymphocytes (5 ϫ 10 5 ) were allowed to migrate in a humidified, 37°C, 5% CO 2 chamber for 2 h to gradients of CCL21. For ␤1 integrin function-blocking studies, primary human T lymphocytes were mixed with 10 g/ml MAB1959 or an isotype control prior to loading in the migration chamber. Following migration, chambers were disassembled, and lymphocytes in the bottom chamber were collected and counted by hemocytometer or fixed in 2% paraformaldehyde and then counted using a hemocytometer. In each assay, migration of cells to serum-free RPMI 1640 medium was considered the "0" migration point, and all migration values were normalized to that point. Assays were performed in duplicate and replicated a minimum of three times.
PLC␥1 shRNA Transfection-Primary human T lymphocytes were transiently transfected with human PLC␥1 shRNA (Ori-Gene TF302445) or control shRNA (Origene TR30015) using LipoD293 (SignaGen) according to the manufacturer's instructions. Briefly, PLC␥1 or control (8 g) shRNAs and LipoD293 were diluted into DMEM. Diluted LipoD293 was added to the diluted shRNA and incubated for 15 min, and the LipoD293-shRNA complexes were added to the lymphocytes. Lymphocytes were incubated with the complexes in complete medium for 72 h at 37°C and 5% CO 2 prior to assay.
␤1 Integrin Activation Assay-Lymphocytes were treated with U0126 or pertussis toxin or transfected with PLC␥1 shRNA, as described previously. Lymphocytes were unstimulated (0 min) or stimulated (0.5-5 min) with 400 nM CCL19 or CCL21. Following stimulation, lymphocytes were submerged in 1 ml of ice-cold 1ϫ PBS. Lymphocytes were labeled with 0.1 g/ml 12G10 on ice for 30 min (or overnight), washed in 1ϫ PBS two times, incubated with FITC-conjugated anti-mouse secondary antibody (Jackson ImmunoResearch Laboratories) for 30 min, and washed in 1ϫ PBS two times. Lymphocytes were fixed in 2% paraformaldehyde and analyzed with the FACSCalibur flow cytometer. All integrin levels were normalized to the mean fluorescence level of integrins found on vehicle control cells (control) for each experiment.
Statistics-All migration and ␤1 integrin activation assays were analyzed by two-tailed, unpaired Student's t test. A p value of Ͻ0.05 was considered statistically significant.

Primary Lymphocytes Differentially Migrate to CCL21 on
Fibronectin-Naive T lymphocytes, which express CCR7, enter the lymph nodes in response to CCL21 found in high endothelial venules. To define the molecular mechanisms that control CCL21-mediated T lymphocyte migration, we used chemotaxis assays. Naive T lymphocytes were induced to migrate over a physiological range of CCL21 concentrations (10 nM to 2 M) or in the absence of CCL21. We found that naive T lymphocytes migrated maximally to 2 M CCL21 on polycarbonate membranes (Fig. 1A). To understand how T lymphocytes migrate via ␤1 integrins, we preincubated the polycarbonate membranes with the ␤1 integrin ligand fibronectin. Lymphocytes were induced to migrate over a wide range of physiological concentrations (10 nM to 2 M) to CCL21, with the greatest migration occurring at 400 nM (Fig. 1B). To confirm that the cells were migrating via ␤1 integrins, we examined the migration in the presence of ␤1 integrin function-blocking antibodies. Under these conditions, the cells failed to migrate across the fibronectin-coated membranes. Isotype controls had no effect on migration (data not shown). These results implicate ␤1 integrins in human T lymphocyte migration to CCL21.
ERK1/2 Is Required for Migration to CCL21-GPCRs can activate ERK1/2 to mediate the migration of human primary macrophages and human osteosarcoma cells (12,14). To examine the role of ERK1/2 in the migration of naive T lymphocytes to CCL21 via their ␤1 integrins, we examined migration to CCL21 in the presence of the MEK inhibitor U0126. Under these conditions, naive T lymphocytes failed to migrate on fibronectin compared with controls treated with an equivalent dilution of vehicle (dimethyl sulfoxide; p ϭ 0.031) ( Fig. 2A). We used Western blotting to confirm that U0126 blocked MEK activation of ERK1/2. As expected, U0126 blocked ERK1/2 phosphorylation in cells treated with CCL21 (Fig. 2B). PLC␥ can mediate the phosphorylation of ERK1/2 (15,16). To examine the contribution of ERK1/2 to activation of PLC␥1, we probed for PLC␥1 phosphorylation. Unexpectedly, we found that phosphorylation of PLC␥1 was also decreased (Fig. 2B). From these results, we concluded that rapid ERK1/2 activation is required for migration to CCL21. Furthermore, PLC␥1 phosphorylation may be mediated by activated ERK1/2 to regulate CCL21-mediated migration.
Cell migration requires regulated integrin adhesion, which is mediated by activation, followed by integrin de-adhesion (17,18). Therefore, we questioned the role of ERK1/2 in ␤1 integrin activation. To this end, we pretreated lymphocytes with U0126 or an identical dilution of the vehicle (dimethyl sulfoxide) and measured the levels of activated ␤1 integrins following activation with CCL21. To examine the state of the ␤1 integrins, lymphocytes were stained with an activation-specific antibody, 12G10 (19). Levels of activated ␤1 integrins were analyzed by flow cytometry. We found that lymphocytes treated with U0126 displayed similar ␤1 integrin activation compared with vehicle controls (control) (Fig.  2C). From these data, we concluded that ␤1 integrin activation in T lymphocytes in response to stimulation of CCR7 with CCL21 is not regulated by ERK1/2 phosphorylation.  Inhibition of ERK1/2 decreases migration of primary naive T lymphocytes to CCL21. Primary human naive T lymphocytes were treated with 1 M U0126 or an equivalent dilution of dimethyl sulfoxide (control) for 90 min. Then, the cells were induced to migrate to CCL21 (10 nM to 2 M) through a 5-m pore membrane that had been preincubated in 10 g/ml fibronectin (n ϭ 3; p ϭ 0.031 for plot of migration of cells pretreated with vehicle compared with lymphocytes pretreated with U0126) (A); stimulated lysates were assayed for phospho-PLC␥, stripped, and re-probed for total PLC␥ or for phospho-ERK1/2, stripped, and reprobed for total ERK1/2 (n ϭ 3) (B); or U0126-treated cells were stimulated with 400 nM CCL21, stained for activated ␤1 integrins, and analyzed by flow cytometry (n ϭ 3; p ϭ 0.41 for integrin activation of control lymphocytes compared with lymphocytes pretreated with U0126) (C).

CCR7/CCL21 Migration via ␤1 Integrins Is Mediated by PLC␥1
DECEMBER 10, 2010 • VOLUME 285 • NUMBER 50 PLC Is Required for Migration on Fibronectin to CCL21-As mentioned previously, PLC␥1 binds to and activates ␤1 integrins (7). Because the loss of ERK1/2 phosphorylation correlated with decreased PLC␥1 phosphorylation and reduced migration to CCL21, we questioned if PLC was necessary to control migration to CCL21. To this end, we used chemotaxis assays in the presence of U73122, a universal PLC inhibitor, or the U73343 control (Fig. 3A). To confirm that U73122 blocked activation of PLC␥ in T lymphocytes in response to CCL21 stimulation, we assayed for phosphorylation of PLC␥1 by Western blotting. Lymphocytes pretreated with U73122 displayed decreased phosphorylation of PLC␥1 compared with the U73343-treated controls (Fig.  3B). Although U73122 blocked all migration, it was unclear why the U73343 control shifted the migratory response of cells. In contrast to the U73343-treated cells, phosphorylation of ERK1/2 was delayed in the U73122-treated cells (Fig. 3B). When T lymphocytes were preincubated with U73122, migration to CCL21 was reduced compared with U73343-treated controls (Fig. 3A). From these results, we concluded that rapid phosphorylation of PLC and/or ERK1/2 mediates the migration of T lymphocytes to CCL21.
PLC␥1 Is Necessary for ␤1 Integrin Migration and Activation-Because U73122 is a universal inhibitor of PLC, to determine the role PLC␥1 in the migration of naive T lymphocytes via ␤1 integrins, we used PLC␥1 shRNA to specifically target and knockdown levels of PLC␥1 in primary human T lymphocytes. Lymphocytes were transiently transfected with PLC␥1 shRNA or control shRNA and assayed for ␤1 integrin activation and migration. Using NIH Image-Quant software, we found that total PLC␥1 levels were reduced by ϳ85% in the shRNA-transfected lymphocytes (Fig. 4A). T lymphocytes that were depleted of PLC␥1 failed to migrate to 400 nM CCL21 on fibronectin-coated membranes, as opposed to T lymphocytes treated with control shRNA (Fig.  4B). In addition, the loss of PLC␥1 led to a failure to activate ␤1 integrins in response to stimulation with 400 nM CCL21, as determined by flow cytometry (Fig. 4C). Because we observed delayed ERK1/2 phosphorylation in response to U73122, we questioned whether this was due to an inhibition of PLC␥1. We observed that the knockdown of PLC␥1 resulted in a trend toward a rapid and sustained increased ERK1/2 phosphorylation compared with controls (Fig. 4, D and E). From these results, we concluded that PLC␥1 is required for rapid ERK1/2 phosphorylation and activation of ␤1 integrins, resulting in the migration of naive T lymphocytes to CCL21 on fibronectin. . PLC mediates migration to CCL21. Primary naive T lymphocytes were preincubated with 2 M U73122 or U73343 for 20 min and induced to migrate to (10 nM to 2 M) CCL21 through 5-m pore fibronectincoated membranes (n ϭ 3; p ϭ 0.0312 for U73122 versus U73343 pretreated cells) (A) or stimulated with 400 nM CCL21 and assayed for phospho-PLC␥, stripped, and re-probed for total PLC␥ or for phospho-ERK1/2, stripped, and re-probed for total ERK1/2 (n ϭ 3) (B). Densitometric analysis was used to analyze the blots. . Depletion of PLC␥1 reduces migration to CCL21. Cells were pretreated with PLC␥1 shRNA and assayed for knockdown (KD) of PLC␥1 (A); induced to migrate through a 5-m pore fibronectin-coated membranes to 400 nM CCL21 (n ϭ 3; p ϭ 0.002 for migration of cells pretreated with PLC␥1 shRNA versus control shRNA) (B); stimulated with 400 nM CCL21 and assayed for levels of activated ␤1 integrins (12G10) by flow cytometry (C); or lysed and assayed for phospho-PLC␥ and re-probed for total PLC␥ or for phospho-ERK1/2 and re-probed for total ERK1/2 (D). E, graphical representation of densitometric measurements of changes in levels of ERK1/2 phosphorylation of blots (n ϭ 3).
G␣ i Mediates Migration to CCL21-To examine the signaling events upstream of PLC␥1, we investigated the role of G proteins in the activation of PLC␥1. CCR7 couples to G␣ i to initiate downstream signaling events and is required for migration to CCL21 (2). Therefore, to examine signaling events more proximal to the membrane that mediate activation of PLC␥1, we pretreated lymphocytes with pertussis toxin. As expected, lymphocytes pretreated with pertussis toxin failed to migrate to any concentration of CCL21 compared with controls (Fig. 5A). Therefore, to determine the role of G␣ i in signaling through CCR7 to mediate activation of PLC␥1 and ERK1/2 in the presence of pertussis toxin, we assayed for changes in phosphorylation of PLC␥1 and ERK1/2 in response to stimulation with 400 nM CCL21. In pertussis toxin-treated lymphocytes, phosphorylation of ERK1/2 and PLC␥1 was reduced following treatment with CCL21 compared with controls (Fig. 5B). These results indicate that CCR7 coupling to G␣ i mediates PLC␥1 and ERK1/2 phosphorylation and ␤1 integrin deactivation, leading to the migration of T lymphocytes on fibronectin.
To determine the extent to which G␣ i affects ␤1 integrin activation in naive T lymphocytes, we pretreated lymphocytes with pertussis toxin, stimulated with 400 nM CCL21, and assayed by flow cytometry for activated integrins on the cell surface. In the presence of pertussis toxin, CCL21 induced the same level of ␤1 integrins as in the CCL21stimulated vehicle control cells (p ϭ 0.365) (Fig. 5C), although cell migration to CCL21 in the presence of pertussis toxin was lost compared with control cells. In addition, there was a significant elevation in the levels of activated ␤1 integrins when lymphocytes were treated with CCL21 in the presence (p ϭ 0.0014) or absence (p ϭ 0.0236) of pertussis toxin compared with unstimulated controls in the presence or absence of pertussis toxin, respectively. There was no significant difference in the levels of activated integrins with or without pertussis toxin prior to treatment with ligand (p ϭ 0.365). Because cell migration requires the control of integrin adhesion (activation) and de-adhesion (deactivation), from these studies, we concluded that signaling through G␣ i leads to cycling of adhesion and de-adhesion states. To test this hypothesis, we examined the activation state of the ␤1 integrins over time (Fig. 5D). We found that although the levels of activated ␤1 integrins varied over time in both pertussis toxintreated cells and control cells, the levels of activation overlapped throughout the time course. These results suggest that G␣ i is important for migration to CCL21, but not for the initial activation of ␤1 integrins in response to CCL21 stimulation.

DISCUSSION
CCR7 is expressed on B lymphocytes, dendritic cells, natural killer cells, T regulatory cells, and naive T lymphocytes and is required for their migration into and within lymph nodes. CCR7 binds to two ligands, CCL19 and CCL21. Although many studies have described CCL19 signaling through CCR7, little is known about how CCL21 signals through CCR7, leading to migration on fibronectin (20 -23).
High endothelial venules and lymph nodes express substantial amounts of fibronectin (1). As lymphocytes enter into and migrate through lymph nodes, they use ␤1 integrins for adhesion and migration (2). Because CCL21 is required for lymph node entry as well as migration within the lymph nodes, it was important to define the CCL21-mediated mechanisms that control ␤1 integrin-mediated migration on fibronectin. In this study, we have provided evidence that CCL21 induces ␤1 integrin activation and promotes migration on fibronectin via PLC␥1 and ERK1/2 phosphorylation.
GPCRs are well known for their ability to activate MAPKs (24,25). MAPKs, which are involved in many cellular processes, including migration (26 -28). Several groups have reported that stimulation of CCR7 by CCL21 mediates phosphorylation of the MAPK ERK1/2 in primary murine T lymphocytes, B lymphocytes, and HEK293 CCR7-transfected cells (11,13,29). However, from these studies, it is difficult to interpret what if any role ERK1/2 plays in the migration of T lymphocytes. In this study, similar to previous studies, we have reported that ERK1/2 is rapidly and transiently phosphorylated in response to CCR7 activation by CCL21 in primary human T lymphocytes. In addition, we found that rapid and transient phosphorylation of ERK1/2 is required for migration to CCL21 via CCR7 on the ␤1 integrin ligand fibronectin.
Transient activation of ERK1/2 has been shown to be important for aortic smooth cell migration (30). We used the MEK inhibitor U0126 to demonstrate that phosphorylation of ERK1/2 is important for T cell migration in response to CCL21. In addition, using PLC␥1 shRNA, we found that ERK1/2 phosphorylation was sustained, which correlated with decreased migration to CCL21. Taken together, these results suggest that rapid and transient phosphorylation of ERK1/2 is important for migration to CCL21 in T lymphocytes.
Stimulation of human epidermoid carcinoma cells with 12(S)-hydroxyeicosatetraenoic acid results in PLC␥1 phosphorylation and downstream activation of ERK1/2 via pertussis toxin-sensitive signaling events (31). In our study, we found that transient phosphorylation of ERK1/2 is lost following reduction of PLC␥1 levels, suggesting that PLC␥1 contributes to the transient phosphorylation of ERK1/2. We determined that not only is ERK1/2 important for migration, but that PLC␥1 is also important for regulating states of ERK1/2 activation. These states of ERK1/2 activation appear to play an important role in ␤1 integrin deactivation.
As mentioned previously, cell migration requires regulated integrin adhesion (activation) followed by de-adhesion (inactivation). Interestingly, we found that G␣ i activation was not necessary for ␤1 integrin activation, even though pertussis toxin inhibited migration to CCL21. Furthermore, we found that pertussis toxin alone did not increase levels of activated ␤1 integrins, although the levels of activated ␤1 integrins increased in the presence of CCL21. Similar to pertussis toxin-treated cells, we found that in the presence of U0126, ␤1 integrins could be activated, yet migration to CCL21 was inhibited. Because pertussis toxin led to reduced levels of phosphorylation of ERK1/2, we speculated that G␣ i and possibly ERK1/2 are not required for the initial activation of ␤1 integrins; however, they are required for de-adhesion (inactivation) of ␤1 integrins. G␣ i is bound to the resting CCR7. Inhibition of G␣ i with pertussis toxin, however, inhibits rapid ERK1/2 phosphorylation and phosphorylation of PLC␥1, but leads to activation of ␤1 integrins. In contrast, a reduction in PLC␥1 levels following treatment with PLC␥1 shRNA blocked integrin activation. One possible explanation is that separate pools of PLC␥1 may be activated by CCR7. In that case, one pool could control ␤1 integrin activation, whereas the second pool would control the phosphorylation state of effectors such as PLC␥1 and ERK1/2. In the absence of G␣ i , ␤1 integrins would be activated; however, the lymphocytes would be unable to regulate PLC␥1 activity to allow for de-adhesion of the ␤1 integrins, and the integrins would remain activated. Alternatively, following treatment with shRNA, other PLCs could be depleted. We concluded that it is the de-phosphorylation of ERK1/2 that is important for ␤1 integrin recycling and migration to CCL21. Studies are currently under way in our laboratory to sort out these hypotheses.
T lymphocytes are crucial mediators of inflammation, autoimmune disorders, allergic disease, and cancer. Migration is not only important for a normal, functioning immune system, but is also important in disease. Therefore, it is important to understand the molecular mechanisms of how T lymphocytes migrate into and throughout lymph nodes to become effector cells to carry out immune responses in the periphery. This study is an important step in further understanding the molecular mechanisms that regulate T lymphocyte migration to CCL21 and provide the basis for pharmacological targets for controlling T cell entry and migration within lymph nodes to regulate the immune response.