Phosphatidylinositol 3-kinase-dependent extracellular calcium influx is essential for CX(3)CR1-mediated activation of the mitogen-activated protein kinase cascade.

Fractalkine, the first member of the CX(3)C chemokine family, induces leukocyte chemotaxis through activation of its high affinity receptor, CX(3)CR1. Like other chemokine receptors, CX(3)CR1 is coupled to a pertussis toxin-sensitive heterotrimeric G(i) protein, which is necessary for rapid rise in the concentration of intracellular calcium. Using a Chinese hamster ovary cell line stably transfected with the CX(3)CR1 receptor, we show that the source of calcium mobilized by fractalkine stimulation is the extracellular pool. Calcium influx is blocked by extracellular calcium chelators, as well as by divalent heavy metals such as Ni(2+), Co(2+), and Cd(2+) without affecting the integrity of intracellular stores. Remarkably, selective phosphoinositide 3-kinase (PI3K) inhibitors, wortmannin and LY294002, abolish the wave extracellular calcium, suggesting that an active PI3K is necessary for this event. The influx of extracellular calcium is in turn required to trigger the activation of the p42/44 mitogen-activated protein/extracellular signal-regulated kinase pathway, but is not necessary for other signals downstream to PI3K, such as phosphorylation of Akt. The potential role of this signaling cascade in fractalkine-mediated chemotaxis is discussed.

The physiological roles of fractalkine are only beginning to be understood. Fractalkine plays a central role in the trafficking of leukocytes in tissues with high blood flow, like the glomerular circuit (6,7). This function is attributed to its unique membrane bound structure, which enables it to form strong adhesive bonds with leukocytes expressing the CX 3 CR1 receptor in this high shear environment (8,9). These adhesive interactions between the membrane-bound fractalkine and cells expressing the fractalkine receptor are independent of CX 3 CR1 signal transduction or integrin function (10). However, fractalkine can also act as a typical chemokine in cell culture assays and in vivo since soluble forms of fractalkine are able to induce the migration of different types of T and natural killer cells through endothelial cells (3,5). A recent study showed that fractalkine is cleaved from membranes of neurons in culture in response to an excitotoxic stimulus (11). In the central nervous system, neuronally derived fractalkine mediates interactions between neurons and microglia upon nerve injury and promotes neuronal cell survival (12)(13)(14). CX 3 CR1, in addition to the CCR5 and CXCR4 receptors, serves as a major co-receptor (along with the CD4) involved in the entry of dual cell tropic HIV-1 strain into target cells (15,16). In agreement with this fact, fractalkine was recently found to inhibit CX 3 CR1 receptor-mediated HIV-1 infection in vitro (15).
Despite its prominent role in inflammation and HIV infection, the signaling pathways triggered by CX 3 CR1 are poorly understood. As with other chemokine receptors, a major signaling consequence of CX 3 CR1 stimulation is a rapid mobilization of calcium. It is not clear, however, how this event is regulated and how elevated intracellular calcium is related to fractalkine-mediated cell migration. Recent studies have pointed at the importance of G␤␥/PI3K␥-dependent signaling cascades in the process of neutrophil, macrophage, and lymphocyte cell migration (17)(18)(19). However, the importance of these pathways in CX 3 CR1 signaling and biological functions has not been addressed. Finally, fractalkine stimulation of CX 3 CR1 has been shown to induce phosphorylation of p42/44 MAPK (20,21), but the architecture of this signaling cascade has yet to be determined. In this study, we have identified a PI3K-dependent signaling cascade, which upon CX 3 CR1 activation modulates extracellular calcium influx and MAPK activation.
Cell Culture-CHO cells were transfected with a linearized vector encoding the CX 3 CR1 receptor generated by PCR amplification of the CX 3 CR1 cDNA from a human brain library. 48 h after transfection, the cells were placed under selection with puromycin (2 g/ml). The most resistant clones were isolated and tested for receptor expression. A highly expressing clone was selected, expanded, and used in all our experiments. Stably transfected CHO-CX 3 CR1 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and puromycin (2 g/ml). Jurkat cells were cultured in RPMI, supplemented with 10% fetal bovine serum, and stimulated with fractalkine as described below.
Receptor Density Determination-CHO cells expressing the human CX 3 CR1 receptor were grown in 15-cm plates for 24 h. The cells were washed three times with warm PBS, and plasma membranes were prepared for radioligand binding assays essentially as described (22). Binding of 125 I-fractalkine (specific activity, 2200 Ci/mmol, PerkinElmer Life Sciences) was performed using 10 g of membrane protein resuspended in 50 mM Tris-HCl (pH 7.4) and 5 mM MgCl 2 according to previously described methods (22). To generate a saturation isotherm, the ligand concentration was varied from 0.2 to 1.2 nM. The reaction (90 min at 25°C) was terminated by rapid filtration, followed by three washes with ice-cold buffer using a cell harvester. Molar excess of cold fractalkine (100 nM) was used to determine nonspecific binding. Binding data from saturation experiments were analyzed using the GraphPad Prism 2.01 program (GraphPad Software, Inc.).
Immunoblotting-For Western blot experiments, cell extracts were prepared using Laemmli sample buffer and subjected to 10% SDS-PAGE and immunoblotted as described (22). Western blot images are representative of at least three independent experiments. In a typical experiment, cells were grown in 12-well plates for 24 h prior to treatment, washed with warm PBS, and incubated in serum-free medium for 2-4 h prior to fractalkine stimulation as indicated in the figures. This treatment was sufficient for detecting phosphorylation of MAPK or Akt over background levels. Incubation with the different inhibitors was initiated 1 h (wortmannin, LY294002, cadmium chloride, nickel chloride, flunarizine, nimodipine, and nifedipine) or 15 min (BAPTA, EGTA) before treatment with fractalkine.
Calcium Mobilization-CHO-CX 3 CR1 cells were washed twice with PBS containing calcium and magnesium. The cells were then resuspended in a total volume of 1 ml of PBS with or without calcium according to the experiment to which two dyes, Fluo-3AM (4 M) and  ) was added individually to all samples just before each test to discriminate between viable and non-viable cells. The cells were then challenged with 50 nM fractalkine, and the response was measured over the course of 90 s. Cells were run at event rates between 100 and 300/s. FlowJo (Treestar Inc.) was used to derive the Fluo-3/Fura Red parameter. The median value for the fluorescence intensity of the ratioed parameter was calculated and displayed over time. All calcium tracings shown in the different figures are representative of at least 3, but in many cases up to 10, independent experiments.
Expression Analysis of T-channel Subunits by RT-PCR-Reverse transcription and PCR were performed to detect the ␣1 subunit transcripts of the T-type Ca 2ϩ channel in CHO-CX 3 CR1 cells. The primers used in the PCR amplification of fragments within the 5Ј coding region of the ␣1G, ␣1H, and ␣1I subunit are described elsewhere (23).

Characterization of CX 3 CR1 Receptor Expression and
Signaling in Stably Transfected CHO Cells-As a first step to examine the signaling pathways stimulated by activation of CX 3 CR1, we established a CHO cell line stably transfected with the human CX 3 CR1 receptor cDNA. We first carried out 125 Ifractalkine binding to cell membrane preparations to verify CX 3 CR1 receptor expression. 125 I-Fractalkine bound in a saturable manner, and Scatchard analysis of the binding data revealed a single population of receptors with a K d of 0.25 nM and a B max of 6.3 pmol/mg protein, representing ϳ250,000 receptors/cell (Fig. 1a). We then tested whether these receptors were functional by determining their ability to mobilize calcium. As shown in Fig. 1b, 50 nM fractalkine induced calcium mobilization, as determined by flow cytometry. The peak of calcium flux was recorded at 3-5 s after agonist exposure and rapidly decreased to background levels in 20 -30 s. These results show that our stably transfected CHO cells express a single class of high affinity CX 3 CR1 receptors for fractalkine ( Fig. 1), which are capable of mobilizing calcium. Because of the importance of PI3K signaling cascades in chemotaxis (17-19, 24 -26), we also examined the phosphorylation of p42/44 MAPK and Akt, two signaling events that typically lie downstream to PI3K activation by GPCRs (27)(28)(29). As indicated in Fig. 1c, incubation of the CHO-CX 3 CR1 cells with fractalkine-induced p44/42 MAPK phosphorylation (upper panel) and Akt phosphorylation (lower panel), peaking at 5 min after agonist addition and rapidly decreasing to background levels 60 min after stimulation. Total MAPK protein levels were unchanged (Fig. 1c, middle panel). No calcium mobilization, MAPK, or Akt phosphorylation was detected upon exposure of untransfected CHO cells to fractalkine (data not shown).  (Fig. 2), even when higher concentrations of fractalkine (500 nM) were used (data not shown). Under these assay conditions, fractalkine does not provoke the release of detectable Ca 2ϩ from intracellular stores, but we could detect the expected increase in cytosolic Ca 2ϩ upon depletion of intracellular stores with thapsigargin in the presence or absence of extracellular calcium (Fig. 2e, and data not shown). In the presence of BAPTA and EGTA, a robust calcium signal was detected when either 5 g/ml ionomycin or 1 M thapsigargin, an inhibitor of endoplasmic reticulum calcium-dependent ATPase, were added to the medium (data not shown). This indicated that incubation of the cells in calcium-free medium or in the presence of calcium chelators, like BAPTA and EGTA, did not affect the integrity of the intracellular calcium stores. This result suggests that the extracellular pool is most likely the source of calcium ions mobilized upon fractalkine stimulation. To evaluate the involvement of a plasma membrane calcium channel, we used divalent cations, which are known general calcium channels antagonists. Low concentrations of NiCl 2 (100 M), CoCl 2 (100 M), and CdCl 2 (100 M) completely blocked the rise in [Ca 2ϩ ] i provoked by fractalkine (Fig. 3, c, d,  and g, respectively), but did not block the release of calcium from the intracellular stores due to exposure to thapsigargin (Fig. 3, d, f, and h). The phospholipase C inhibitor U73122 (5 M) had no effect on calcium influx provoked by fractalkine (Fig. 3b). Because fractalkine binds to CX 3 CR1 (Fig. 1a) and induces Akt phosphorylation in the absence of Ca 2ϩ (see text below and Fig. 7b), the effects of extracellular Ca 2ϩ depletion and the effects of heavy metals on Ca 2ϩ influx are not simply due to a dependence of fractalkine receptor binding on Ca 2ϩ . Taken together these results suggest that the rise in [Ca 2ϩ ] i is strictly dependent on the influx of extracellular calcium through plasma membrane channels and does not involve Ca 2ϩ release from intracellular stores stimulated by activation of a U73122-sensitive phospholipase C␤-mediated mechanism.

Fractalkine-mediated CX 3 CR1 Receptor Activation Causes Influx of Calcium Ions from the Extracellular
CX 3 CR1 Receptor-induced Calcium in Influx Requires a Flunarizine/Amiloride-sensitive Calcium Channel-Voltage-gated calcium channels have been divided into several subtypes based on their conductance and sensitivities to voltage. This class of channels is most typically expressed in excitable cells. However, several reports have suggested the presence of L-and T-type channels in non-excitable cells as well, such as lymphocytes and fibroblasts (30 -33). We used selective Ca 2ϩ channel antagonists to examine whether L-or T-type channels mediate the influx caused by exposure to fractalkine. Two specific Ltype channel blockers, nifedipine (10 M) and nimodipine (10 M) failed to block fractalkine-induced calcium influx (Figs. 4,  e and d, respectively), ruling out the involvement of an L-type channel. T-type channels are harder to characterize pharmacologically due to the lack of truly specific antagonists. However, a few compounds, among them flunarizine and amiloride, have selective inhibitory effects on T-type channels when used at low concentrations (23,34,35). Fig. 4 (b and c) shows that flunarizine (10 M) and amiloride (100 M) effectively blocked calcium influx by fractalkine at concentrations considered selective for T-type channels (23, 34, 35), but they did not inhibit calcium release from the intracellular stores provoked by exposure to 1 M thapsigargin (data not shown). Of these two compounds, amiloride can also inhibit Na ϩ /H ϩ exchange systems, but higher concentrations (Ͼ1 mM) are typically required (34). The calcium influx signal stimulated by fractalkine was insensitive to TTX (Fig. 5b), excluding the possibility that a TTX-sensitive Na ϩ channel would be admitting sufficient Ca 2ϩ to cause an increase in [Ca 2ϩ ] i . We then asked whether this non-excitable cell line would trigger calcium influx upon depolarization with 30 mM KCl. This treatment, however, failed to induce calcium influx (Fig. 5c), suggesting the absence of functional voltage-activated channels in the CHO-CX 3 CR1 cell line. Activation of a potassium channel can lead to hyperpolarization and consequent influx through voltage-insensitive Ca 2ϩ channels (36), but such mechanism cannot explain our data because triethanolamine (30 mM), an inhibitor of K ϩ currents had no effect of fractalkine Ca 2ϩ influx (Fig. 5d). Another possibility would be an initial activation of T-type channels, perhaps present in the cells at very low levels, followed by Ca 2ϩ -induced Ca 2ϩ release from intracellular stores. This would be in agreement with the very low expression of T-type channels found in ovary tissue (37). We thus tested the expres- sion of three recently identified T-type calcium channels ␣1 subunits, namely ␣1G, ␣1H, and ␣1I, in our CHO-CX 3 CR1 cells using RT-PCR with primers specific for the 5Ј of their coding regions (23), and using as positive control a sample from a human brain cDNA library. We did not detect expression of any of these subunits in these cell lines (data not shown). Altogether, we interpret these results to indicate the involvement of a plasma membrane calcium channel sensitive to divalent cations, as well as flunarizine and amiloride. However, despite the reported selectivity of flunarizine and amiloride at the concentrations we used, our data are inconsistent with these being voltage-activated T-type channels.
PI3K Is Required for CX 3 CR1 Receptor-induced Calcium Influx and MAPK Activation-Reports in the literature suggest that GPCRs and receptor tyrosine kinases can both activate voltage-dependent as well as receptor-activated calcium channels through direct interaction of the G protein with the channels as well as through the action of PI3K (38 -40). Because CX 3 CR1 receptor activation results in PI3K stimulation (Fig.  1c), and calcium influx via a plasma membrane Ca 2ϩ channel (Figs. 2 and 3), we asked whether these two signaling events triggered by the CX 3 CR1 receptor were part of the same signaling cascade. For this purpose, we preincubated the cells with two chemically different and selective inhibitors of PI3K, wortmannin (200 nM) and LY294002 (25 M) before stimulating with fractalkine and measuring calcium influx. Both PI3K inhibitors significantly attenuated the fractalkine-mediated calcium influx (Fig. 6, b and c), without affecting the integrity of intracellular stores, as evidenced by a strong calcium signal obtained upon incubation of the cells with thapsigargin in the presence of wortmannin and LY294002 (data not shown). In addition, preincubation of the cells with concentrations as low as 5 M LY294002 and 50 nM wortmannin was sufficient to reduce the phosphorylated MAPK signal to background levels (Fig. 7a), and they inhibited, as expected, the activation of Akt (not shown). Taken together, these results indicate that PI3K is required for the stimulation of calcium influx and MAPK.
Calcium Influx Is Required for MAPK but Not for Akt Phosphorylation-Because elevation of [Ca 2ϩ ] i is necessary for activation of the MAPK pathway by some GPCR agonists, we examined whether this was also the case for fractalkine-induced signaling to MAPK. As it is shown in Fig. 7b, MAPK phosphorylation was significantly reduced by EGTA (1 mM), BAPTA (1 M), as well as by the divalent cations Ni 2ϩ (2 mM) and Cd 2ϩ (100 M). In experiments using calcium channel antagonists, flunarizine (10 M) effectively blocked MAPK activation, whereas the L-type blockers nifedipine (10 M) and nimodipine (10 M) had no significant effect (Fig. 7b). In contrast, the calcium channel inhibitors did not affect the induction of Akt phosphorylation at Ser 473 (Fig. 7b), which depends upon PI3K activation. Consistently, increase in [Ca 2ϩ ] i due to the depletion of intracellular stores by thapsigargin (1 M), induced MAPK but not Akt phosphorylation (Fig. 7c). In turn, inhibition of calcium entry blocks the MAPK pathway, but inactivation of MEK by 20 M PD98059 has no effect on CX 3 CR1 receptor-mediated calcium influx (Fig. 6d). These results locate PI3K upstream of calcium channel activation, which in turn is necessary for subsequent activation of the MAPK pathway. DISCUSSION The effects of chemoattractants such as formylmethionyleucylphenylalanine, interleukin 8, and C5a on neutrophil migration as well as those of regulated on activation normal T cell expressed and secreted, macrophage inflammatory protein-5, macrophage-derived chemokine, and stromal cell-derived factor-1 on macrophage migration have underscored the importance of G␤␥/PI3K␥-dependent cascades in the process of chemotaxis (17)(18)(19). We thus hypothesized that signaling cascades emanating from G␤␥ and PI3K could also play an essential role in CX 3 CR1-mediated chemotaxis. The increase in cytosolic calcium concentration is a typical signaling readout of chemokine receptor activation. However, the significance of fast calcium mobilization with respect to cell migration is still being debated. To begin analyzing these questions in the context of CX 3 CR1-mediated signaling cascades, we established a CHO cell line expressing the CX 3 CR1 receptor (Fig. 1). Stimulation with 10 nM fractalkine induced robust signaling responses such as fast phosphorylation of MAPK and Akt (Fig.  1c), as well as a fast increase in cytosolic calcium concentrations (Fig. 1b). The increase in [Ca 2ϩ ] i could be the result of G␤␥ activation, and subsequent phospholipase C␤-mediated release of calcium from intracellular stores. Alternatively, the increase in [Ca 2ϩ ] i could be caused by influx of extracellular calcium through plasma membrane channels (41). Our results demonstrate that fractalkine-mediated elevation of [Ca 2ϩ ] i is caused by a potent influx of extracellular calcium (Figs. 2 and  3). This conclusion is based on the observations that the rise [Ca 2ϩ ] i is: (a) completely dependent on the presence of extracellular calcium (Fig. 2b); (b) blocked by non-permeable calcium chelators such as BAPTA (20 M) and EGTA (5 mM) (Fig. 2, c  and d), under conditions that do not affect the integrity of the intracellular stores; (c) inhibited by general Ca 2ϩ channel blockers such as Ni 2ϩ , Cd 2ϩ , and Co 2ϩ cations (all 100 M) (Fig.  3, c-e); and (d) not affected by inhibition of phospholipase C with U71322 (5 M) (Fig. 3b). Extracellular calcium mobilization in response to fractalkine is thus mediated directly by plasma membrane Ca 2ϩ channels, and does not seem to depend on prior release of Ca 2ϩ from intracellular stores.
The calcium influx generated by activation of CX 3 CR1 is blocked by Pertussis toxin (Ref. 5, and data not shown), which indicates the requirement of G o/i coupling. G␤␥ subunits can bind directly and inhibit N-, P-, and Q-type voltage-gated channels predominantly present in excitable cells (42)(43)(44), but typically absent in non-excitable cells such as fibroblasts and lymphocytes. G␤␥ dimers can also activate receptor-operated channels and voltage-operated L-type channels (38,40). For example, GPCR agonists such as angiotensin II, noradrenalin, and GnRH can induce the opening of L-type channels (40,45). However, two different 1,4-dihydropiridine type/L-type channel inhibitors, nimodipine and nifedipine, did not affect fractalkine-induced calcium influx at high concentrations (Fig. 4). In contrast, the inhibitors flunarizine and amiloride significantly reduced the CX 3 CR1-mediated Ca 2ϩ influx (Fig. 4), when used at concentrations considered to be specific for T-type channels (23,34,35). However, two lines of evidence stand against the notion that these effects are truly mediated by T-type channels. First, depolarization with 30 mM KCl did not induce calcium influx (Fig. 5c), suggesting the absence of voltage-gated channels in the CHO-CX 3 CR1 cells; and second, we could not detect expression of ␣1G, ␣1H, or ␣1I subtype T-type channel mRNA in CHO-CX 3 CR1 cells by RT-PCR. Thus, the identity of the channel mediating fractalkine calcium influx remains to be determined. Our current hypothesis is that the rapid rise in [Ca 2ϩ ] i is mediated by yet uncharacterized receptor-operated channels with unreported sensitivity to those compounds. The use of flunarizine and amiloride could actually help in identifying these channels. It is intriguing in this context that mibefradil, another relatively selective T-type channel blocker, can produce potent immunosuppression by inhibiting transmigration of CD4ϩ and CD8ϩ T cells through allogeneic endothelium (46). It would be also interesting to extend the analysis of the mechanisms of fractalkine-induced Ca 2ϩ influx to excitable neurons, where CX 3 CR1 is abundantly expressed and plays a role in fractalkine-mediated neuronal survival (12,14).
Earlier reports have indicated a regulatory role for PI3K signals on calcium channels, particularly of the L-and N-types (39,40). This prompted us to test whether fractalkine-mediated Ca 2ϩ influx was also mediated by PI3K. Indeed, two chemically different and selective PI3K inhibitors, wortmannin and LY294002, effectively reduced Ca 2ϩ influx, whereas the specific MEK inhibitor PD98059 had no effect (Fig. 6). Both PI3K inhibitors also blocked fractalkine-induced activation of MAPK (Fig. 7), indicating that PI3K lies upstream of both Ca 2ϩ and MAPK signaling events. In turn, inhibition of Ca 2ϩ signaling by the extracellular Ca 2ϩ chelators BAPTA and EGTA, and by Ca 2ϩ channel-blocking cations Ni 2ϩ and Cd 2ϩ inhibited MAPK phosphorylation (Fig. 7b). Similar to the effects on Ca 2ϩ influx, flunarizine also reduced MAPK phosphorylation, but this effect was not observed with the L-type antagonists, nimodipine and nifedipine (Fig. 7b). Attenuation of Ca 2ϩ entry, however, had no effect on the phosphorylation of Akt at Ser 473 , which also requires activation of PI3K (Fig. 7b). Consistently, the rise in [Ca 2ϩ ] i by depletion of intracellular stores did not stimulate Akt phosphorylation (Fig. 7c). Taken together, these results delineate a signaling cascade involving the activation of PI3K. PI3K-derived signals then bifurcate in two separate effector pathways, one regulating the influx of extracellular calcium through a plasma membrane channel, and the other activating Akt and possibly other downstream effectors in a Ca 2ϩ -independent manner. Elevation of [Ca 2ϩ ] i triggered by fractalkine stimulation is in turn required for MAPK pathway activation. Current efforts are under way to dissect more thoroughly the FIG. 7. PI3K and calcium influx are required for CX 3 CR1 receptor-mediated MAPK phosphorylation. a, dose-dependent inhibition of MAPK phosphorylation by wortmannin and LY294002. CHO-CX 3 CR1 cells were preincubated with wortmannin and LY294002 for 60 min before stimulation for 5 min with 10 nM fractalkine. Control lanes contain extracts from non-induced cells. b, effect of calcium channels inhibitors on fractalkine-mediated MAPK phosphorylation (upper panel) and Akt phosphorylation (lower panel). Cells were preincubated for 15 min with BAPTA (1 M) and EGTA (1 mM), or for 1 h with CdCl 2 (100 M), NiCl 2 (2 mM), flunarizine (10 M), nifedipine (10 M), and nimodipine (10 M) before stimulation with fractalkine as indicated above. Extracts were prepared for immunoblotting with antibodies against phosphorylated (upper panel) and total MAPK (lower panel) as described under "Experimental Procedures." Control lanes contain extracts from non-induced cells. c, time course of thapsigargin (Tg) effects on MAPK and Akt phosphorylation. Ca 2ϩ -activated MAPK pathway. This important cascade involves three common sequential steps, namely activation of PI3K, elevated [Ca 2ϩ ] i , and activation of MAPK.
Fractalkine is thus a remarkable dual action molecule, regulating cell-cell interactions (10) and cell migration. Our preliminary studies suggest that a similar signaling cascade involving PI3K, [Ca 2ϩ ] i , and MAPK is essential for fractalkineinduced chemotaxis of Jurkat T cells. Thus, signaling pathways diverging downstream of PI3K appear to play multiple and important roles in CX 3 CR1 function such as chemotaxis and cell survival (our preliminary results and Refs. 12 and 14, respectively). As PI3K and Akt may work to promote the polarization of signaling proteins required for triggering chemotaxis (47), the question of how PI3K signals regulate [Ca 2ϩ ] i , and consequently MAPK deserves further investigation. Interestingly, a recent report suggests that phosphatidylinositol 3,4,5-trisphosphate, the main product of PI3K, induces calcium entry through T cell plasma membrane channels by a novel uncharacterized mechanism (48). The Ca 2ϩ -mediated activation of the MAPK cascade may in turn play a role in the remodeling of the cytoskeleton required for CX 3 CR1-induced cell movement, perhaps by enhancing myosin light chain kinase activity and the phosphorylation of downstream targets involved in cell motility (49). Future studies designed to uncover the different components of this pathway could shed more light on the molecular events triggering the chemotactic response to fractalkine during inflammation.