Roles for phosphoinositide 3-kinases, Bruton's tyrosine kinase, and Jun kinases in B lymphocyte chemotaxis and homing


  • Saida Ortolano,

    1. B Cell Molecular Immunology Section, Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA
    Search for more papers by this author
  • Il-Young Hwang,

    1. B Cell Molecular Immunology Section, Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA
    Search for more papers by this author
  • Sang-Bae Han,

    1. B Cell Molecular Immunology Section, Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA
    Search for more papers by this author
  • John H. Kehrl

    Corresponding author
    1. B Cell Molecular Immunology Section, Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA
    • Building 10, Rm. 11B08, NIAID, NIH, Bethesda, MD 20892, USA, Fax: +1-301-402–0070
    Search for more papers by this author


B lymphocyte chemokine receptors signal to downstream effectors by activating heterotrimeric G proteins. However, many of these effectors remain unknown and the known ones often have ill-defined roles in B cell trafficking. Here we report that pharmacological inhibitors of phosphoinositide 3-kinases (wortmannin, WMN), Bruton's tyrosine kinase (LFM-A13), and Jun kinases (SP600125) all significantly impair CXCL12-induced mouse B cell chemotaxis and that of a human B lymphoma cell line. Examination of two CXCR4-induced signaling pathways revealed that LFM-A13 and WMN blocked Akt activation, while SP600125 and WMN blocked JNK activation. Each of the inhibitors impaired the homing of transferred B cells to peripheral lymph nodes. Intravital imaging of control and inhibitor-treated mouse B cells in the inguinal lymph node high endothelial venules (HEV) demonstrated a 17%, 35%, and 60% reduction in the number of firmly adherent B cells with LFM-A13, SP600125, and WMN, respectively. These results implicate chemokine receptor mediated activation of phosphoinositide 3-kinases in the firm adhesion of mouse B cells within peripheral lymph node HEV, while Bruton's tyrosine kinase and JNK activation are less important and more likely needed during B cell transmigration through the endothelium and/or trafficking into the lymph node parenchyma.


Bruton's tyrosine kinase


intracellular calcium




5, 6-[(4-chloromethyl) benzoyl] aminotetramethylrhodamine


high endothelial venules


: α-cyano-β-hydroxy-β-methyl-N-(2,5-dibromophenyl)propenamide


phosphorylated protein kinase B


phosphorylated JNK


phospholipase C




While the importance of chemokine receptors for lymphocyte trafficking and for the appropriate positioning of lymphoid cells is well established 1, 2, the signaling pathways activated by lymphocyte chemokine receptors and the physiological roles of these signaling pathways remain poorly understood. Furthermore, those studies that have examined chemokine receptor signaling pathways have largely focused on T rather than B lymphocytes, and in those instances where both B and T cells have been studied, clear differences have often emerged.

The entrance of both B and T cells into lymph nodes and Peyer's patches depends upon chemokine receptor-triggered G protein activation. Inhibition of Gi activation blocks the firm adhesion of lymphocytes to high endothelial venules (HEV) by inhibiting integrin activation 3, 4. The mechanism by which Gi activation leads to integrin activation requires the integration of at least three independent signaling pathways controlled by PI3K, Rap1, and RhoA 57. Highlighting the difference between B and T cells, B cells require DOCK2 for integrin activation, while T cells do not 8. A recent study indicated that WMN treatment impaired the entrance of mouse B cells into Peyer's patches and the splenic white pulp, but did not statistically reduce entrance into the inguinal and mesenteric lymph nodes 9. Following firm adhesion to HEV, lymphocytes transmigrate through the endothelium and enter the parenchyma of lymph nodes or Peyer's patches. CXCR5-mediated signaling recruits B cells into B cell follicles and CCR7 signaling facilitates B and T cell interactions at the interface of the B cell follicle with the T cell zone 10, 11.

An early consequence of chemokine receptor signaling is an alteration of lymphocyte morphology. Lymphocytes acquire a polarized shape with a leading edge, where chemokine receptors cluster, and a trailing edge or uropod, where adhesion molecules such as CD44, ICAM-1, and ICAM-3 accumulate 12. Helping to control some of these morphological changes are three small GTPases, Cdc42, Rac, and RhoA 13. Each has been assigned a distinctive role: Cdc42 in the extension of filopodia, Rac in formation of lamellipodia, and RhoA in cell retraction and release of the uropod from substrates 14. However, linking the activation of heterotrimeric G proteins to small GTPase activation has been complicated by the bewildering number of exchange factors for these molecules, although Vav, an exchange factor for Rac, has been recently implicated downstream of CXCR4 signaling in human peripheral blood lymphocytes 15. In addition, roles for PI3K and Tec kinases in the CCXCL12-induced Cdc42 and Rac activation in T lymphocytes has been ascribed 16.

Downstream of the small GTPase are a number of signaling pathways linked to chemotaxis and cell motility including protein kinases such as the mitogen-activated protein kinases (MAPK), Paks, and p160ROCK. While chemokine signaling in lymphocytes activates the MAPK, extracellular signal-regulated kinase (Erk), there is little evidence that its activation is required for chemotaxis or lymphocyte migration 17. A recent study supports a role for the MAPK p38 in CXCL12-induced chemotaxis of acute lymphocytic leukemia cells and perhaps in normal lymphocytes 18. Despite being essential for the migration of numerous cell types 19, 20 no functional role for JNK activation in lymphocyte chemotaxis or migration has been documented.

The motility of B cells within lymph nodes depends, at least in part, upon heterotrimeric G protein-mediated signaling since B cells from Gnai2–/– mice move poorly within the inguinal lymph node, and treatment of wild-type mice with pertussis toxin reduces B cell motility within the inguinal lymph node 21. Furthermore, the addition of CXCL12 increases the motility of B cells purified from mouse spleens incorporated into collagen matrices 21. The downstream signaling pathways that control lymphocyte motility likely include those involved in triggering lymphocyte polarity, integrin activation, actin-cytoskeletal rearrangements, and de-adhesion.

In this study, we have examined a variety of inhibitors of signaling molecules potentially involved in B cell chemotaxis. We focused on three, WMN because of the known involvement of PI3K in lymphocyte chemotaxis 8, 9, LFM-A13 because of the importance of Bruton's tyrosine kinase (Btk) in B lymphocyte function and recent demonstration of its role in neutrophil chemotaxis 22, and JNK because of its known role in the migration of other cell types 19, 20.


Identification of inhibitors of B lymphocyte chemotaxis and/or migration

In preliminary experiments we screened seven inhibitors of signaling molecules either known or suspected to be involved in lymphocyte chemotaxis and/or migration in a standard filter-based chemotaxis assay using purified mouse splenic B cells and CXCL12. Of the seven inhibitors, five of them significantly impaired chemotaxis. The most active ones were the Btk inhibitor LFM-A13, the PI3K inhibitor WMN, the JNK inhibitor SP600125, the p38 inhibitor SB03580, and the protein kinase C inhibitor calphostin. The two inhibitors that were less effective were the JAK kinase inhibitor AG490 and the rho-associated protein kinase (p160ROCK) inhibitor Y27632 (data not shown).

We elected to focus on three of the inhibitors WMN, LFM-A13, and SP600125. While optimal concentrations for WMN treatment of murine B cells are established 9, 23, there is much less experience in using LFM-A13 and SP600125. We found that both LFM-A13 and SP600125 reduced B cell chemotaxis to CXCL12 in a dose-dependent fashion (Fig. 1a). In addition, each reduced the spontaneous migration noted in the absence of added chemokine. A comparison between optimal concentrations of the three inhibitors in a mouse chemotaxis assay to CXCL12 is shown in Fig. 1b.

Figure 1.

Dose-response curves to LFM-A13, SP600125, and WMN. (a) Dose response chemotaxis assay with LFM-A13 and SP600125. Splenic B cells were pre-incubated with increasing concentrations of LFM-A13 or SP600125. The results are shown as the percent total migration in the absence or presence of CXCL12 (50 ng/mL). Data are mean ± SD of triplicate determination. (b). Chemotaxis of mouse B cells pre-treated with optimal concentrations of LFM-A13, SP600125, or WMN. Splenic B cells were pre-treated with LFM-A13 (100 μg/mL), SP600125 (50 μM), or WMN (50 nM) and subjected to a chemotaxis assay using CXCL12 (50 ng/mL). Results are mean ± SD, (*p<0.05 versus control). The dotted line represents spontaneous migration of control cells in the absence of chemokine. (c) Effect of increasing concentrations of LFM-A13, SP600125, and WMN on chemotatic response to CXCL12. A human B cell lymphoma cell line (BJAB) was pre-incubated with increasing concentrations of LFM-A13, WMN, or SP600125 and subjected to a chemotaxis assay with CXCL12 (100 ng/mL). Results are shown as specific migration (CXCL12-induced – spontaneous migration).

To provide further evidence that these inhibitors affected signaling pathways important in B cells chemotaxis, we examined the three inhibitors on the chemotactic response of human B lymphoma line (BJAB) to CXCL12 (Fig. 1c). We required a slightly lower concentration of LFM-A13 and SP600125 to inhibit BJAB chemotaxis than we had observed with the mouse B cells. To inhibit mouse B cell chemotaxis with WMN, we used a 50 nM concentration, while we required 80 nM to similarly inhibit the B cell lymphoma cell line.

To verify that the three inhibitors did not alter the expression of cell surface receptors known to be involved in chemotaxis and/or cell migration, we examined the expression of CXCR4, CXCR5, CD44, CD62, LFA-1, and α4β7 integrin on mouse B cells after incubation with three inhibitors or DMSO as a control. None of the inhibitors significantly altered the expression of these molecules in a manner that would impair their functional responses (Table 1).

Table 1. Effect of various inhibitors on the surface expression of chemokine receptors and cell adhesion molecules
Mean fluorescence% cellsMean fluorescence% cellsMean fluorescence% cellsMean fluorescence% cells

Effect of WMN, SP600125, and LFM-A13 on CXCL12-induced increase in [Ca2+]i and B cell motility

Next, we examined the motility of the mouse B cells in collagen matrices either in the presence or absence of CXCL12. We tracked cells either by manually tracking individual cells in sequential images acquired using an inverted microscope every 30 s for 10 min (data not shown) or by differentially labeling control and inhibitor treated cells and acquiring images every 30 s for 10 min using a spinning disk confocal microscope. The fluorescent images were analyzed by software capable of tracking individual cells. Both methods gave similar results. Each of the inhibitors slightly reduced the basal velocity of the B cells compared to control B cells (Fig. 2). CXCL12 treatment enhanced the velocity of the control B cells by approximately 40%. The inhibitors also reduced the velocity of the B cells treated with CXCL12; however, only LFM-A13 significantly narrowed the gap between the non-CXCL12-treated and CXCL12-treated B cells (Fig. 2). We also examined the effects of the same three inhibitors on the motility of the human B cell lymphoma cell line. Each inhibitor minimally affected the basal motility of the cell line. Addition of CXCL12 doubled the velocity and, in contrast to the primary mouse B cells where only LFM-A13 had a significant effect, each of the inhibitors reduced the velocity of the CXCL12-stimulated lymphoma cells compared to control cells (Fig. 2).

Figure 2.

Effect of LFM-A13, SP600125, and WMN on B cell motility. Effect of inhibitors on B cell motility in collagen matrices. Mouse B cells or BJAB cells were pre-incubated with media containing LFM-A13, SP600125, WMN, or control media for 90 min prior to differentially labeling with CMMTR (inhibitor treated) or CMFDA (control). The cells were then incorporated into a three-dimensional collagen matrix in glass-bottom culture dishes. After 30 min, the cells motilities were analyzed by confocal microscopy. During image acquisition cells were maintained at 37ºC. Image stacks containing five images were captured every 30 s for 10 min either in presence or absence of 50 ng/mL CXCL12. The image data were processed and cells velocities calculated. Results are mean ± SD, (*p<0.05 versus control).

Next, we looked at each of the inhibitors on CXCL12-induced increases in [Ca2+]i . We treated mouse B cells with each of the three inhibitors or DMSO as a control, and then loaded the cells with a calcium indicator dye and exposed the cells to CXCL12. We measured changes in [Ca2+]i over time. Both WMN and LFM-A13 significantly reduced the CXCL12-induced increases in [Ca2+]i normally observed following CXCL12 exposure; in contrast, SP600125 had little effect. Individual calcium tracings from control cells and inhibitor treated cells as well as the peak calcium response related to the control cells are shown (Fig. 3).

Figure 3.

Effect of inhibitors on CXCL12-induced increase in [Ca2+]i. Mouse B cells were incubated for 90 min with various inhibitors and plated in poly D-lysine-coated plates and loaded with the calcium indicator prior to the addition of CXCL12 (100 ng/mL). Two tracings for SP600125 and LFM-A13 are shown, whereas single tracings are shown for WMN and the control. The calcium flux peak was measured over 3 min, and data were analyzed with SOFT max Pro. The tracings from control and inhibitor-treated cell are indicated. The graph below compares the peak calcium response of control to inhibitor-treated cells (*p<0.05 versus control).

Effect of WMN, LFM-A13, and SP600125 on CXCL12-induced increases in phosphorlyated Akt and phosphorylated JNK

In these experiments, we stimulated control and experimental groups with CXCL12 for various durations and prepared cell lysates at various time points following the addition of CXCL12. As a measure of PI3K activation, we analyzed the cell lysates for the presence of phosphorylated protein kinase B/Akt (pAkt) 23, 24. PI3K catalyze the formation of phosphatidylinositol 3,4 bisphosphate [PtdIns(3,4)P2] and PtdIns 3,4,5-trisphophate [PtdIns(3,4,5)P3], which allows the recruitment and activation of effector molecules such as Btk and Akt, which contain a PH domain specific for these 3′phosphoinositides 25. As expected WMN significantly reduced the presence of pAkt in cell lysates of CXCL12-stimulated B cells as compared to control cells, but in addition LFM-A13 significantly reduced Akt activation as well (Fig. 4a).

Figure 4.

Effect of various inhibitors on CXCL12-induced activation of Akt and JNK in mouse B cells. (a) CXCL12-induced Akt activation. Mouse B cells were pre-treated with LFM-A13, WMN, or with the DMSO control prior to stimulation with CXCL12 (100 ng/mL) for the indicated durations. The cells lysates were analyzed for pAkt or Akt levels by immunoblotting. (b) CXCL12 induced JNK activation. Similar experiment to (a), but the mouse B cells were pre-treated with each of the three inhibitors prior to CXCL12 stimulation. Cell lysates were examined for expression of pJNK and JNK by immunoblotting.

Next, we examined JNK activation in splenic B cells stimulated with CXCL12. We found that exposure to CXCL12-induced JNK activation, as assessed by the appearance of phosphorylated JNK (pJNK) in the cell lysates at 1, 5, and 10 min after stimulation. The addition of SP600125, as expected, significantly reduced JNK activation. In contrast, LFM-A13 had little effect on JNK activation; however, we found a significant reduction in JNK activation following treatment with WMN (Fig. 4b). In addition, treatment of a human B cell lymphoma cell line with WMN markedly reduced the basal levels of pAkt and pJNK, and markedly inhibited CXCL12-induced increases of both (data not shown).

WMN, LFM-A13, and SP600125 inhibit the entrance of lymphocytes into inguinal and popliteal lymph node

To determine the effects of these inhibitors in a more physiological assay, we performed lymph node entrance assays by differentially labeling inhibitor-treated and control B cells, and transferring the B cells intravenously to a recipient animal. After 2 h, we collected the spleen, inguinal lymph nodes, popliteal lymph nodes, and peripheral blood. We then assessed the percentage of control and inhibitor-treated cells among the cells prepared from each of these locations by flow cytometry. We have extensive experience using these two dyes and have had observed no effect of either dye on the homing of lymphocytes 21. We found that LFM-A13 pre-treatment did not affect the recovery of cells from either the spleen or the blood, but that it significantly reduced the recovery of transferred cells found in the inguinal lymph nodes and the popliteal lymph nodes (Fig. 5). Treatment with SP600125 also had no effect on the recovery of cells from the spleen and blood, and also reduced the recovery from the inguinal and popliteal lymph nodes (Fig. 5). WMN treatment reduced the number of recovered cells from the spleen, had no effect on the recovery of cells from the blood, and modestly reduced the number recovered from inguinal and popliteal lymph nodes (Fig. 5).

Figure 5.

LFM-A13, SP600125, and WMN decrease B lymphocyte homing. Mouse B cells were treated with LFM-A13 (100 μg/mL), SP600125 (50 μM), WMN (50 nM), or control media prior to differential labeling with CMFDA (inhibitor treated) or CMMTR (control). An equal number of inhibitor-treated and control cells were transferred intravenously to recipient mice. At 2.5 h after transfer, cells from blood, spleen, inguinal lymph nodes, and popliteal lymph nodes of recipient mice were analyzed by flow cytometry. Shown are the relative number of transferred B cells found in blood, spleen, inguinal lymph nodes, and popliteal lymph nodes (*p<0.05 versus control).

WMN and to a lesser extent SP600125 reduce the fraction of B cells sticking to HEV in the inguinal lymph node

To test whether the reduced entrance of inhibitor-treated B cells into lymph nodes was secondary to reduced interactions with HEV, we used intravital microscopy to analyze the behavior of fluorescently labeled inhibitor treated or control B cells in the inguinal lymph node microcirculation of mice. The hierarchical branching of venules within peripheral lymph nodes has five distinct venular orders, with orders 3–5 corresponding to paracortical HEV 26, 27. We focused on these higher order venules, and collected a series of movies between 30 min and 120 min after transfer. Fig. 6 shows representative images 30 min after transfer revealing a predominance of control B cells (green) versus either the SP600125, LFM-A13, or WMN treated B cells (red) in the peripheral venules. To provide a quantitative assessment of B cell adhesion in the inguinal lymph node HEV, we determined the percentage of rolling cells and their sticking efficiency for each inhibitor. The lower right panel in Fig. 6 shows that the rolling fraction of lymphocytes was slightly, although not significantly, increased by pre-treatment of the B cells with LFM-A13 or SP600125 when compared to control cells; however, the fraction of rolling cells was significantly increased by prior treatment with WMN. Of those B cell rolling in the HEV, a fraction firmly adhere to the HEV. The sticking fraction of B cells was most impacted by prior treatment with WMN, less so by SP600125, and least affected by LFM-A13. The reductions were 60%, 35%, and 17%, respectively.

Figure 6.

LFM-A13, SP600125, and WMN reduce B cell sticking to HEV in the inguinal lymph node. Mouse B cells were treated with LFM-A13 (100 μg/mL), SP600125 (50 μM), WMN (50 nM), or control media prior to differential labeling with CMFDA (inhibitor treated) or CMMTR (control). An equal number of inhibitor-treated and control cells were transferred intravenously to recipient mice. Intravital imaging began as soon as possible after transfer (usually within 15 min). Images were recorded every 1 s for 2-min intervals. Representative images from approximately 30 min after transfer for inhibitor-treated cells (red arrows) and control cells (green arrows) are shown. The total flux of cells was determined by counting all cells that passed through each venule. The rolling fraction was calculated by dividing the flux of rolling cells by the total flux. The sticking fraction was defined as the percentage of rolling cells that became firmly adherent for >30 s. Data are mean ± SD (35 venules, 2 mice, *p<0.05) for each condition.


Pharmacological inhibitors of PI3K, Tec kinases, and JNK substantially reduced CXCL12-induced chemotaxis of mouse B cells and a human B cell line, and impaired select signaling pathways. These studies provided the basis for an assessment of the inhibitors in two more physiological assays. Pre-treatment with each inhibitor reduced the trafficking of mouse B cells to lymph nodes in recipient mice and pre-treatment with the PI3K inhibitor significantly reduced the sticking of transferred B cells to HEV in the inguinal lymph node of recipient mice.

PI3K have emerged as important mediators in chemoattractant signaling particularly in studies of the model organism Dictyostelium discoideum28, 29. By analogy the class IB PI3K, p110γ, a downstream effector of Gβγ signaling, was an obvious candidate to mediate chemokine receptor signaling in mammalian cells. Indeed, mice lacking p110γ have neutrophils, monocytes, and T lymphocytes that are hypo-responsive to chemokines 3032. Surprisingly, B cells from these mice are not impaired in chemotaxis or homing assays 8, 9. A recent study has suggested a role for the class I PI3K p110δ in CXCL13-induced B cell chemotaxis, although not in response to CCL21, CCL19, or CXCL12. The same study showed that WMN treatment impaired B lymphocyte homing to Peyer's patches and splenic white pulp, but that it did not statistically reduce homing to inguinal lymph nodes 9. In our study, we used a slightly lower concentration of WMN, although we pre-incubated for a longer period. Our results indicate that WMN pre-treatment significantly impairs splenic B cell chemotaxis to CXCL12, impairs inguinal and popliteal lymph node homing, and impairs B cell/HEV interactions. We also showed that WMN significantly reduced CXCL12-mediated chemotaxis and B cell motility of human B lymphocyte cell line (similar results were found with human peripheral blood B cells; S. Ortolano, unpublished data). Together our data indicate a role for a PI3K isoenzyme in B cell/HEV interactions and B cell chemotaxis. Additional experiments with specific inhibitors of individual PI3K isoenzymes should help delineate their relative importance and avoid the problems of compensation inherent in the analysis of gene-targeted mice.

WMN treatment of mouse B cells also impaired CXCL12-induced Akt and JNK activation, and reduced CXCL12-induced increases in [Ca2+]i. The reduction in Akt activation was expected, while the decreased JNK activation and impaired increase in [Ca2+]i were not. JNK and PI3K activation have not been frequently linked, although lipopolysaccharide-induced JNK activation in neutrophils is sensitive to WMN as well as IgE-induced JNK activation in mast cells 33, 34. EGF-induced JNK activation in HeLa cells is also sensitive, while tumor necrosis factor-induced activation is not 35. PI3K have also been linked to Ca2+ signaling in other cell types via an inhibition of the activation of L-type calcium channels and by affecting intracellular Ca2+ release 36, 37.

Although not previously implicated in lymphocyte chemotaxis, JNK is a downstream target of several signaling pathways that regulate cell migration and chemotaxis, including Rac and cdc42 38. JNK phosphorylates paxillin, which has been shown to be necessary for the migration of fish keratocytes, a cell type known for their rapid actin-dependent motility 19. The role of paxillin in lymphocyte migration has received little attention. Another study showed that SP600125 blocks IL-8-induced migration of rat basophilic leukemia cells 39. Our studies indicate that the chemotaxis of mouse B cells and a human B cell lymphoma line to CXCL12 is impaired by SP600125, as is the homing of mouse B cells to peripheral lymph nodes. Inhibition of JNK activation did not impair CXCL12-induced increases in [Ca2+]i and had a modest effect on B cell motility in collagen. Analysis of the interaction of SP600125-treated B cells with HEV in the inguinal lymph node indicated that the inhibition of JNK activation reduced the fraction of cells sticking, in part accounting for the reduce homing. Presumably, SP600125 also interfered with transmigration and entrance into the lymph node parenchyma to fully account for its reduction in homing.

There is increasing evidence for a role of Tec kinase family members in leukocyte chemotaxis. Chemokine receptor signaling can trigger Tec kinase activation and translocation to the plasma membrane 16. Both Gβγ subunits and G12 have been directly implicated in Btk activation 40, 41. In addition, CXCR4 signaling has been shown to activate Src kinases 42, which have a prominent role in the activation of Tec kinases. In T cells, the expression of a dominant-negative form of Itk impairs CXCL12-induced cell polarization and migration, as well as Cdc42 and Rac activation 16. In neutrophils, LFM-A13 treatment inhibits fMet-Leu-Phe-stimulated increases in cell adhesion, chemotaxis, and phospholipase D activity. In the same study, LFM-A13 impaired fMet-Leu-Phe-induced translocation of Rac, RhoA, and Btk to the plasma membrane 22. Finally, the analysis of Btk–/–Tec–/– mice has revealed a markedly disrupted splenic architecture with few clearly defined follicular structures, a phenotype consistent with defective chemokine receptor signaling 43. However, despite the interest in Btk in B cells, its role in B cell chemotaxis has not received much attention. We found that LFM-A13 inhibited the chemotactic response of mouse B cells and a human B cell line to CXCL12. In addition, it inhibited CXCR4-triggered increases in [Ca2+]i and Akt activation. Since LFM-A13 inhibits both Tec and Btk, these results cannot be solely attributed to the inhibition of Btk.

Btk is known to amplify the rise in [Ca2+]i triggered by B cell antigen receptor cross-linking by facilitating the activation of phospholipase C (PLC)γ and enhancing the synthesis of PtdIns(4,5)P2, which is required for its upstream activator PI3K and downstream target PLCγ2 44, 45. However, chemokine receptor-mediated increases in [Ca2+]i are triggered by the release of Gβγ subunits, which activates PLCβ. One possibility is that chemokine receptor-triggered Btk activation enhances the synthesis of PtdIns(4,5)P2, which augments PI3K activity and serves as a substrate for PLCβ. Another possibility is that, since several G protein-coupled receptors directly interact and stimulate the activity of PLCγ1 46, CXCR4 signaling could recruit PLCγ in a similar fashion creating a target for Btk.

There is some precedent for Btk having a role in Akt activation. An interaction between Btk and Akt, which enhanced Akt activation, has been observed following B cell antigen receptor cross-linking 47. Similar to what we found, LFM-A13 reduced Akt activation following the stimulation of neutrophils with fMet-Leu-Phe 22. The inhibition of Akt activation by LFM-A13 and by WMN suggests that Akt activation may contribute to B cell chemotaxis. A recent study that examined human B cell chemotaxis to CCL21 found that pre-exposure to an Akt inhibitor reduced the response by approximately 50% 48.

Of the three inhibitors studied here, LFM-A13 had the least effect on the fraction of sticking cell in the inguinal lymph node HEV. While Btk has been implicated in integrin activation following stimulation of B cells via the B cell antigen receptor, our studies suggest that Btk activation does not impact B cell adhesion to inguinal lymph node HEV. One caveat in this conclusion is that in our study the cells are removed from the inhibitor prior to tail vein injection, and the intravital imaging is performed approximately 30 min later allowing for the potential wash-out of the inhibitor. However, the in vivo B cell homing assay indicates that there is sufficient impairment to reduce homing to peripheral lymph nodes. Based on the reduction of chemoattractant-induced increases of activated Cdc42, Rac, and RhoA found by interfering with Tec kinases in T cells and neutrophils 16, 22, in B cells Btk may modulate the activities of the exchange factors that trigger the activation of one or more these small GTPases.

This study provides an impetus for further studies of the role of Btk and JNK in B cell chemotaxis, and continued studies of the role of PI3K isoenzymes particularly in B cell adhesion. Study of the downstream signaling pathways activated by Btk and/or Tec during B cell chemotaxis will focus upon the small GTPase exchange factors prominently expressed in B cells. Whether JNK phosphorylates paxillin in B lymphocytes and regulates B cell polarity and migration, as has been described in fish keratocytes, will also be examined.

Materials and methods

Mice and reagents

Female mice C57BL/6, age 6–8 weeks, were obtained from The Jackson Laboratory (Bar Harbor, ME) and housed according to guidelines of Institutional Animal Care Committee (NIH). Chemicals LFM-A13, SP600125, SB203580, AG490, Y27632, calphostin C, and WMN were purchased from EMD Bioscience, (La Jolla, CA) and aliquoted in stock solutions of 1 mg/mL in DMSO. Recombinant mouse CXCL12 was purchased from R&D System (Minneapolis, MN). Anti-mouse antibodies CD11c, Gr-1, CD4,CD8 were obtained from BD Pharmingen (San Diego, CA). Cell trackers CMTMR {5,6-[(4-chloromethyl)benzoyl]aminotetramethylrhodamine} and CMFDA (5-chloromethylfluoresceindiacetate) were purchased from Molecular Probes (Eugene, OR). Antibodies against Akt, pAkt (Ser473), JNK I and II, pJNK I and II (Thr183/Tyr185) were purchased from Cell Signaling (Beverly, MA).

B cell purification

B cell were isolated from mouse spleen by negative depletion, as previously described 49, using biotinylated antibodies to CD11c, Gr-1, CD4, and CD8 (BD Pharmingen) and streptavidin-conjugated Dynebeads (Dynal Inc., Oslo, Norway). Cell purity of greater than 95% was verified by flow cytometry.

Inhibitors treatment

Freshly isolated splenic B cells, resuspended in RPMI 1640 medium (Invitrogen Inc., Carlsbad, CA), were incubated prior to each experiments, for 1.5 h with either LFM-A13, SP600125, WMN, AG490, calphostin, SB203580, or Y27632 or control 0.2% DMSO at 37°C and 5% CO2 atmosphere.


Chemotaxis assays were performed using transwell chambers (Costar, Corning, NY) as previously described 49. Cells (105/well), pre-treated with one of the inhibitors or DMSO for control samples, were added to the upper part of a 24-well transwell plate with 5-μm inserts. The lower chambers were filled with RPMI 1640 media containing inhibitor or solvent, 10% FBS, and CXCL12 (50 ng/mL). After incubation for 3 h at 37ºC and 5% CO2, cells in the lower wells were collected and counted by flow cytometry. Total cell input and spontaneous migration in absence of chemokine were also measured. Results are shown as the percentage of total or specific migration. Total migration is calculated by dividing the migrated cells by the number of input cells. The specific migration is calculated by dividing the number of migrated cells in the presence of chemokine minus the number of migrated cells in the absence of chemokine by the input cell number.

Cell motility

Mouse B cells were stained with cell tracker CMTMR (red) for inhibitor-treated samples, or CMFDA (green) for control cells, following the manufacturer instructions (2 μM, 15 min at 37°C). Fluorescent cells in equal number for each condition, were cultured, as previously described 50 in a three-dimensional collagen matrix (1.67 mg/mL type I collagen pH 7.0, Angiotech Biomaterials, Palo Alto, CA) in RPMI 1640 plus 10% FBS, and allowed to polymerize at 37°C and 5% CO2 atmosphere for 30 min in glass-bottom culture dishes (MatTeck Corporation, Ashland, MA). Cell motility was analyzed by confocal microscopy using a Zeiss inverted microscope equipped with Perkin-Elmer Ultraview, argon/krypton laser and Orca-ErII CCD camera (Hamamatsu, Japan). During the image acquisition the cells were maintained at 37ºC and 5% CO2 atmosphere using an environmental controller system (Carl Zeiss). Images were captured every 30 s for 10 min either in presence or absence of CXCL12 (50 ng/mL) using a 20× Plan-Apochromat objective. Images were processed and cells velocity calculated using Imaris 4.04 software (Bitplane AG, Zurich, Switzerland).

Intracellular Ca2+ levels

Inhibitor-treated or control cells were seeded at a density of 20 000 cells in 100 μL loading medium (RPMI 1640 plus 10% FBS) into poly-D-lysine-coated 96-well black-wall, clear-bottom microtiter plates (Nalge Nunc, Rochester, NY). An equal volume of assay loading buffer (FLIPR Calcium 3 assay kit, Molecular Devices, Sunnyvale, CA) in Hanks’ balanced salt solution supplemented with 20 mM HEPES and 2 mM probenecid was added. Cells were incubated for 1 h at 37 C before adding chemokine, and then the calcium flux peak was measured on a FlexStation (Molecular Devices). The data were analyzed with SOFT max Pro (Molecular Devices).

Western blot

Either control or inhibitor-treated B cells were stimulated with CXCL12 (100 ng/mL) for varying durations. Cell lysates were prepared as previously described 21. The detergent-insoluble materials were removed by centrifugation for 10 min at 4°C. Equal amounts of proteins from each sample were fractionated by 10% SDS-PAGE and transferred to pure nitrocellulose. Membranes were blocked with 5% BSA in Tween 20 plus TBS (TTBS) for 1 h and then incubated with an appropriate dilution of the primary antibody in 5% BSA in TTBS for 2 h. The blots were incubated with biotinylated antibodies for 1 h, and further incubated with streptavidin-conjugated to horseradish peroxidase for 1 h. The signal was detected by enhanced chemiluminescence according to the manufacturer's instructions (Amersham Pharmacia Biotech, Piscataway, NJ).

Entrance assay

Splenic B cells previously treated with one of the three inhibitors were stained (15 min 37ºC, 5% CO2) with CMTMR and control cells with CMFDA. The cells were mixed at a 1:1 ratio (107 of each) and injected intravenously into the tail vein of a recipient mouse. After 2.5 h, the animal was killed. Lymph nodes (popliteal and inguinal) and spleen were harvested and blood was collected by cardiac puncture. Erythrocytes were removed with Tris-NH4Cl and lymphocytes resuspended in PBS containing 1% BSA. The number of fluorescent cells in each collected sample was calculated by flow cytometry using a FACSCalibur (BD Bioscience). Data were analyzed using FlowJo software (Tree star Inc., San Carlos, CA).

Intravital imaging

Inguinal lymph node microcirculation was imaged by intravital microscopy as previously described with minor modifications 21, 26. Equal numbers of inhibitor-treated and control cells were stained as described for the entrance assay and injected into the tail vein of a recipient mouse and immediately imaged. An Axiovert 135 inverted microscope (Carl Zeiss) with 20× Plan-Apochromat objective was used to collect time-lapse images every 1 s for 2 min. Different movies were acquired up to 1.5 h after injection. Imaris 4.04 (Bitplane AG) was used for image processing. The rolling and sticking fractions were calculated as previously described 51. Cells in 35 venules in each of two different mice for each condition were counted. Total flux of cell was determined by determining the rolling, sticking and non-interacting cells passing through each venule for a determined period of time. The rolling fraction was calculated by dividing the flux of rolling cells by the total flux. The sticking fraction was defined as percentage of rolling cells firmly adherent for more than 30 s.

Flow cytometry

Inhibitor-treated and control B cells (105 per sample) were stained with anti-mouse L-selectin, LFA-1 and CD44 FITC-conjugated antibodies and anti-mouse CXCR4, CXCR5 and α4β7 integrin PE-conjugated antibodies (Molecular Probes), according to manufacturer instructions. The percentage of cells expressing each marker, as well as mean fluorescence intensity were acquired by flow cytometry (FACSCalibur) and further analyzed with FlowJo software.


Results for in vitro assays are expressed as mean of triplicate samples in three independent experiments. For in vivo homing assays three mice in three independent experiments were used for each condition. Standard deviation and p value (Student's t-test) were calculated using Microsoft Excel software. Values were considered significantly different for p value <0.05.


The authors would like to thank Ms. Mary Rust for her editorial assistance, Dr. Ning-Na Huang for her assistance with the imaging experiments, and Dr. Anthony Fauci for his continued support.


  1. 1


  2. 2


  3. 3


  4. 4


  5. 5


  6. 6