Knock-out Mice Reveal the Contributions of P2Y and P2X Receptors to Nucleotide-induced Ca2+ Signaling in Macrophages*

Immune cell function is modulated by changes in extracellular nucleotide levels. Here we used reverse transcription-PCR analyses, single cell Ca2+ imaging, and knock-out mice to define the receptors mediating nucleotide-induced Ca2+ signaling in resident peritoneal macrophages. In Ca2+-free buffer, the potent (K0.5 <1 μm) stimulatory effect of UTP (or ATP) on endoplasmic reticulum (ER) Ca2+ release was abolished in cells isolated from P2Y2/P2Y4 double knock-out mice. Moreover, \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{P}2\mathrm{Y}_{4}^{0{/}-}\) \end{document}, but not \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{P}2\mathrm{Y}_{2}^{-{/}-}\) \end{document}, macrophages responded to UTP. In \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{P}2\mathrm{Y}_{2}^{-{/}-}\) \end{document} macrophages, we could elicit Ca2+ responses to “pure” P2X receptor activation by applying ATP in buffer containing Ca2+. Purified UDP and ADP were ineffective agonists, although modest UDP-induced Ca2+ responses could be elicited in macrophages after “activation” with lipopolysaccharide and interferon-γ. Notably, in Ca2+-free buffer, UTP-induced Ca2+ transients decayed within 1 min, and there was no response to repeated agonist challenge. Measurements of ER [Ca2+] with mag-fluo-4 showed that ER Ca2+ stores were depleted under these conditions. When extracellular Ca2+ was available, ER Ca2+ stores refilled, but Ca2+ increased to only ∼40% of the initial value upon repeated UTP challenge. This apparent receptor desensitization persisted in GRK2+/- and GRK6-/- macrophages and after inhibition of candidate kinases protein kinase C and calmodulin-dependent kinase II. Initial challenge with UTP also reduced Ca2+ mobilization by complement component C5a (and vice versa). In conclusion, homologous receptor desensitization is not the major mechanism that rapidly dampens Ca2+ signaling mediated by P2Y2, the sole Gq-coupled receptor for UTP or ATP in macrophages. UDP responsiveness (P2Y6 receptor expression) increases following macrophage activation.

Antigen-presenting cells such as monocytes, macrophages, and dendritic cells express two families of nucleotide receptors as follows: ATP-gated cation channels (P2X receptors) and P2Y receptors, a subset of the G protein-coupled receptor superfamily (1). In mouse, ATP and UTP are equipotent agonists at the G q -coupled receptors P2Y 2 and P2Y 4 , whereas human P2Y 4 is selectively activated by UTP and competitively antagonized by ATP (2). Moreover, in human, ATP additionally activates the dual G s -and G q -coupled P2Y 11 receptor, absent in the mouse genome (3). The concentration of ATP or UTP in the local extracellular milieu is estimated to be around 10 nM (4,5), which may be just sufficient to evoke local inositol 1,4,5trisphosphate (IP 3 ) 2 -induced Ca 2ϩ puffs (6). However, mechanical stress, cellular injury, inflammation, degranulation of mast cells, and other factors may increase ATP and UTP to levels sufficient to evoke larger and longer lasting global Ca 2ϩ signals (5,7,8). ATP is also released as a cotransmitter from the sympathetic nervous system, a potentially important neuroimmune interaction (9,10).
Local increases in ATP or UTP levels are transient because of both diffusion and the activity of ecto-nucleotidases such as CD39 (NTPDase 1), which catalyzes the sequential hydrolysis of ATP and UTP to their respective monophosphates (8,11). A subset of P2Y receptors is selectively activated by nucleotide diphosphates. In particular, P2Y 1 and P2Y 12 receptors are preferentially activated by ADP and play important roles in platelet function (12). UTP degradation by CD39 yields transiently the intermediate UDP, which is a specific agonist for P2Y 6 receptors (13). During prolonged agonist stimulation, the response of P2Y and P2X receptors is typically switched off within minutes (14 -16). In the case of G protein-coupled receptors, the activity of the activated G protein is terminated by GTPase-activating proteins, which catalyze the hydrolysis of GTP bound to the ␣-subunit (17). Agonist-induced "desensitization" is thought to involve phosphorylation of the C terminus (or intracellular loops) by G protein-coupled receptor kinases (GRKs) or second messenger-induced protein kinase C (PKC). Phosphorylation by GRK promotes the rapid binding of ␤-arrestin, which blocks further G protein activation (18,19).
Because of the lack of specific agonists and antagonists, it is difficult to assign unequivocally P2Y and P2X receptor subtypes to a particular cell type. In this study, by using P2Y 2 -and/or P2Y 4 -deficient mice, as well as RT-PCR analyses, we show that P2Y 2 is the dominant receptor in resident peritoneal macrophages. Furthermore, we dissected the interplay of P2Y 2 receptors, P2X receptors, Ca 2ϩ release-activated Ca 2ϩ channels, and Ca 2ϩ -ATPases (Ca 2ϩ pumps) in nucleotide-induced Ca 2ϩ signaling. Finally, we assessed the role of PKC, calmodulin-dependent kinase II (CaMKII), GRK2, and GRK6 in the rapid desensitization of P2Y 2 receptors using inhibitors and GRKdeficient mice.
Isolation of Peritoneal Macrophages-Mice were killed by cervical dislocation, and resident peritoneal cells were harvested by lavage with 10 ml of ice-cold Hanks' physiological salt solution. After centrifugation, cells were resuspended in RPMI medium containing 10% heat-inactivated fetal calf serum, 100 units/ml penicillin, and 100 g/ml streptomycin. Cells were seeded onto glass coverslips and incubated at 37°C in air with 5% CO 2 . After 2 h, nonadherent cells were removed by washing the coverslip. In selected experiments, macrophages were activated by incubation for Ͼ24 h in RPMI medium (with 10% calf serum) containing 100 ng/ml lipopolysaccharide and 100 units/ml interferon-␥.
Fluorescence Microscopy-Differential interference contrast and fluorescence images of live peritoneal macrophages, typically 12-14 m diameter, were acquired using an Olympus IX71 microscope equipped with a software-controlled Sensi-CamQE CCD camera (Chromaphor, Duisburg, Germany). Cells were labeled with the fluorescent nucleic acid stain Hoechst 33342 (Invitrogen) and imaged using a ϫ100 (1.4 numerical aperture) objective lens and immersion oil. Endoplasmic Reticulum [Ca 2ϩ ] Measurements-To measure ER [Ca 2ϩ ] ([Ca 2ϩ ] ER ), the low affinity fluorescent Ca 2ϩ indicator mag-fluo-4 was selectively loaded into the ER. Cells were incubated at 37°C with 5 M mag-fluo-4/AM (Molecular Probes), washed several times, and incubated overnight at 4°C before use. Probenecid was omitted from the Hanks' buffered solution to promote leak of cytosolic mag-fluo-4. Fluo-3 or mag-fluo-4 loaded macrophages were excited at 488 nm via a monochromator, whereas fluorescence was detected at 530 Ϯ 15 nm. Only one cell per coverslip was used for experiments, and moreover, only the first response to a given nucleotide concentration was used to establish concentration-response relations. The fluorescence signals were normalized with respect to the resting fluorescence intensity (F 0 ) and expressed as F/F 0 .
Materials-Chemicals were obtained from Sigma unless stated otherwise. Solutions of UDP or ADP were purified from contaminating nucleotide triphosphates by incubation with 50 units/ml hexokinase and 20 mM glucose for at least 2 h at room temperature. Recombinant mouse interferon-␥ was obtained from PeproTech (London, UK), and recombinant mouse complement component C5a was purchased from R & D Systems (Wiesbaden-Nordenstadt, Germany).

P2Y Receptor Signaling-
To determine the effects of extracellular nucleotides on ER Ca 2ϩ release in macrophages, Ca 2ϩfree buffer was used to obviate the potentially confounding effects of store-operated Ca 2ϩ entry and Ca 2ϩ influx via P2X receptors. When single WT macrophages were superfused with UTP in Ca 2ϩ -free buffer, a concentration-dependent increase in peak cytosolic [Ca 2ϩ ] ([Ca 2ϩ ] i ) was evoked (Fig. 1A); UTP increased peak [Ca 2ϩ ] i with a K 0.5 value of 0.7 M. Activation of either P2Y 2 or P2Y 4 receptors could be responsible for the transient Ca 2ϩ response to UTP in Ca 2ϩ -free buffer. Both of these receptor subtypes have recently been reported to be expressed in mouse macrophages (23). To explore the relative roles of P2Y 2 and P2Y 4 receptors, we first challenged macrophages isolated from P2Y 2 and P2Y 4 knock-out mice with UTP. Concentrations of UTP between 0.1 and 100 M, or as high as 250 M (n ϭ 7; not shown), had no effect on [Ca 2ϩ ] i in macrophages isolated from P2Y 2 /P2Y 4 double knock-out mice (Fig. 1B), indi-cating that P2Y 2 or P2Y 4 are the only possible G q -coupled receptors mediating UTP-induced Ca 2ϩ signaling. P2Y 2 /P2Y 4 double knock-out mice exhibited no obvious phenotype (20), and macrophages isolated from these animals were indistinguishable from WT cells (Fig. 2, A and B). Moreover, similar to WT cells, macrophages isolated from double knock-out mice expressed mRNA specific for P2Y 1 and P2Y 6 ( Fig. 2C), and P2Y 2 , but not P2Y 4 , could be detected in WT cells. Mouse whole brain RNA was used as positive control for the P2Y 4 -specific primer (not shown). Both WT and double knockout macrophages expressed several ATP-gated, nonselective cation channels as follows: P2X 1 , P2X 4 , and P2X 7 receptors (Fig. 2D). The upper band in the P2X 4 lanes was not identified.
To assess the relative contributions of P2Y 2 and P2Y 4 receptors, we compared the effects of UTP on Ca 2ϩ signaling in macrophages isolated from P2Y 2 Ϫ/Ϫ or P2Y 4 0/Ϫ mice. Similar to WT cells, application of 10 M UTP (in Ca 2ϩ -free buffer) elicited a transient and oscillatory increase in cytosolic Ca 2ϩ in macrophages isolated from P2Y 4 0/Ϫ mice, as shown in Fig. 3A. However, as illustrated by the example in Fig. 3B, macrophages isolated from P2Y 2 -deficient mice did not respond to UTP, but these cells responded to ATP, provided that Ca 2ϩ was present in the buffer (Fig. 3B). The concentration-response relations for macrophages isolated from P2Y 4 0/Ϫ and P2Y 2 Ϫ/Ϫ mice are presented in Fig. 3, C and D. Thus, consistent with the RT-PCR data, the concentration-response data reveal that P2Y 2 is the sole G q -coupled receptor for UTP in resident macrophages.
UTP is sequentially degraded to UDP and UMP extracellularly, a reaction catalyzed by the surface membrane enzyme  Note the bean-shaped morphology of the nuclei, typical for macrophages. B, overlay of fluorescence and differential interference contrast images of the macrophages shown in A. The scale bar is 10 m. C and D, RT-PCR analyses of RNA obtained from purified macrophages isolated from P2Y 2 /P2Y 4 double knock-out and WT mice. P2Y 1 and P2Y 6 receptors were detected in both DKO and WT macrophages, and P2Y 2 , but not P2Y 4 , could be detected in WT cells (C). In both cases, P2X 1 , P2X 4 , and P2X 7 receptors were detected (D). All PCR products were confirmed by sequencing. The upper band shown in the P2X 4 lanes was not identified. NOVEMBER 17, 2006 • VOLUME 281 • NUMBER 46 CD39. Because we detected mRNA for P2Y 6 in both WT and double knock-out macrophages, we expected to observe ER Ca 2ϩ release induced by UDP. In preliminary experiments, 100 M UDP (97% pure by high performance liquid chromatography) consistently induced ER Ca 2ϩ release. However, after contaminating UTP (or ATP) was scavenged with hexokinase (24), 10 of 11 WT macrophages did not respond at all to 100 M UDP, whereas all cells responded to ATP, used as positive control (Fig. 4A). Note that when extracellular Ca 2ϩ was reintroduced after stimulating cells with high ATP (or UTP; not shown) concentrations in Ca 2ϩ -free buffer, an increase in intracellular Ca 2ϩ , consistent with store-operated Ca 2ϩ entry (25), was always observed (Fig. 4A). Moreover, similar to WT, 12 of 17 P2Y 2 /P2Y 4 -deficient macrophages did not respond to 100 M UDP, and the response in the remaining five cells was characterized by a small, single Ca 2ϩ transient (Fig. 4B). Taken together, the [UDP]-response relations for macrophages from both WT and P2Y 2 /P2Y 4 double knock-out mice were essentially flat in the 1-100 M range (Fig. 4C). These data reveal that, functionally, P2Y 6 is weakly expressed in resident macrophages. However, when P2Y 2 /P2Y 4 -deficient macrophages were activated for 48 -72 h with lipopolysaccharide and interferon-␥, responsiveness increased, such that most cells (9 of 11) produced a small oscillatory Ca 2ϩ response to 100 M UDP (peak F/F 0 3.0 Ϯ 0.2; not shown).

P2Y and P2X Receptors in Macrophages
To explore the potential role of P2Y 1 receptors, we challenged macrophages with ADP under Ca 2ϩ -free conditions. Application of 100 M ADP (purified with hexokinase) had no effect on [Ca 2ϩ ] i in 9 of 11 WT cells, whereas ATP produced a strong positive control response (Fig. 5A). Concentration-response relations for ADP and ATP in WT macrophages are overlaid in Fig. 5B. Similar to WT, ADP also had little or no effect in macrophages isolated from P2Y 2 /P2Y 4 double knock-out mice, and no response to ATP was observed (Fig. 5C).
P2X Receptor Signaling-Although we could not elicit a Ca 2ϩ response to UTP in macrophages isolated from P2Y 2 -deficient (P2Y 2 Ϫ/Ϫ or double knock-out) mice, we could observe a small transient response to ATP provided that Ca 2ϩ was present in the buffer (Fig. 6). Thus, targeted disruption of the P2Y 2 gene reveals the Ca 2ϩ response to pure P2X receptor activation. The Ca 2ϩ response elicited by ATP was small and diminished upon repeated application of agonist (Fig. 6, A and B). The concentration-response relation in the 0.1-250 M ATP range is shown in Fig. 6C (data from P2Y 2 Ϫ/Ϫ and double knock-out mice were combined). Note that in RT-PCR analyses we could detect P2X 1 , P2X 4 , and P2X 7 in macrophages isolated from both WT and P2Y 2 /P2Y 4 double knock-out mice (see Fig. 2D). For comparison, the pure P2Y receptor response to ATP is overlaid in Fig. 6. It can be seen that the P2Y receptor-induced Ca 2ϩ response greatly dominates over the P2X receptor response in the 0.1-10 M ATP range.

GRK6 Is Not Essential for Rapid Desensitization of P2Y 2
Receptors-When extracellular Ca 2ϩ was available, application of 100 M UTP typically produced a rapid increase in [Ca 2ϩ ] i followed by fast and slow components of decline (Fig. 7A), the latter of which was absent in Ca 2ϩ -free buffer (for example, see Fig. 4A) and can be attributed to store-operated Ca 2ϩ entry. Furthermore, the Ca 2ϩ response to repeated UTP challenge was greatly diminished (Fig. 7A), suggesting that the initial challenge may have desensitized the receptor. GRKs have been implicated in the desensitization mechanism of various G protein-coupled receptors, and to test whether GRK6 is involved in the rapid desensitization of P2Y 2 receptors, we isolated macrophages from GRK6 Ϫ/Ϫ mice. Compared with WT cells, there was no enhancement of the second UTP-induced Ca 2ϩ response in GRK6-deficient macrophages, even when cells were pretreated with inhibitors of the candidate regulatory kinases PKC (staurosporine) and CaMKII (KN93) (Fig. 7B). RT-PCR analyses of RNA extracted from purified macrophages indicated that GRK2 and GRK5 were also expressed, but only weak signals for GRK3 and GRK4 were detected (Fig. 7C). We also found that apparent desensitization, assayed by repeated challenges with 100 M UTP, was not reduced in macrophages isolated from GRK2 ϩ/Ϫ mice, which express 50% protein compared with WT (26). Thus, the experiments with WT, GRK2 ϩ/Ϫ , and GRK6 Ϫ/Ϫ macrophages summarized in Fig. 7D (see also supplemental Fig. 2) suggest that the kinases GRK2, GRK6, PKC, and CaMKII are not necessary for the rapid dampening of P2Y 2 receptor signaling in macrophages.

Stimulation of P2Y 2 Receptors Decreases the Subsequent Response to a Non-P2Y G q -coupled Receptor-If the decreased
Ca 2ϩ response to repeated UTP application is because of homologous P2Y 2 receptor desensitization, then the response to a non-P2Y G q -coupled receptor should be unaffected following initial UTP challenge. To test this possibility, we used complement factor C5a as a second agonist, the receptor of which has been shown using WT versus G q/15 Ϫ/Ϫ mice to be coupled to G q/15 (27), whereas P2Y receptors are coupled to G q/11 . When a macrophage was first challenged with 100 M UTP followed by 100 nM C5a, the second Ca 2ϩ response was decreased (Fig. 8A) and vice versa (Fig. 8B), indicating that heterologous mechanisms, such as phosphatidylinositol 4,5-bisphosphate depletion, are contributing to the decayed Ca 2ϩ responses to repeated UTP challenges. On average, the peak of the second Ca 2ϩ response was ϳ40% of the initial response when UTP was applied twice, whereas the second response was ϳ60% (relative to the first response) when different agonists were used (Fig. 8C).
Another potentially important factor determining the magnitude of the second agonist-induced Ca 2ϩ response is the availability of ER Ca 2ϩ . Indeed, there was no response at all to repeated UTP challenge when experiments were performed in Ca 2ϩ -free buffer (Fig. 8C), an unphysiological condition that promotes Ca 2ϩ extrusion via the plasma membrane Ca 2ϩ -ATPase (Ca 2ϩ pump).   NOVEMBER 17, 2006 • VOLUME 281 • NUMBER 46

P2Y and P2X Receptors in Macrophages
Desensitization-Mag-fluo-4 was selectively loaded into the ER lumen to measure [Ca 2ϩ ] ER . When 100 M UTP was applied to macrophages, a decrease in [Ca 2ϩ ] ER was observed (Fig. 9A), which on average recovered 77.3 Ϯ 4.5% (n ϭ 6). Compared with fluo-3, mag-fluo-4 was much more susceptible to photobleaching, and thus the extent of recovery was probably underestimated in these experiments. In Ca 2ϩ -free buffer, the recovery of [Ca 2ϩ ] ER after transient application of UTP was less than 15% (n ϭ 5), as shown in Fig. 9B. However, the ER lumen was rapidly refilled with Ca 2ϩ after switching to buffer containing 1.3 mM Ca 2ϩ (Fig. 9B).

DISCUSSION
Calcium is an important regulatory ion inside cells, and in immune cells it has been implicated in diverse functions, FIGURE 7. Apparent P2Y 2 receptor desensitization persists in GRK2 ؉/؊ macrophages and in GRK6 ؊/؊ macrophages subjected to PKC and CaMKII inhibition. A, typical response of a WT macrophage to repeated UTP challenges in the continued presence of extracellular Ca 2ϩ . Note that the second Ca 2ϩ response to UTP is considerably reduced. B, apparent receptor desensitization (decay of the Ca 2ϩ response with repeated application of UTP) is not obviated in macrophages isolated from GRK6 Ϫ/Ϫ mice and pretreated with inhibitors of PKC (staurosporine (Staurosp.)) and CaMKII (KN93). C, RT-PCR analyses of RNA obtained from purified WT macrophages. The GRKs GRK2, GRK5, and GRK6 were strongly expressed, whereas only weak signals for GRK3 and GRK4 could be detected. D, summary of experiments showing that, compared with control conditions, the second Ca 2ϩ response to repeated agonist challenge (UTP to UTP) was not augmented in macrophages obtained from GRK2 ϩ/Ϫ or GRK6 Ϫ/Ϫ mice, and no difference was observed with inhibitors of the candidate regulatory kinases PKC and CaMKII.
The number of single cells tested in each group is indicated in parentheses.

FIGURE 8. Heterologous mechanisms contribute to apparent P2Y 2 receptor desensitization.
A and B, the Ca 2ϩ response to complement factor C5a is reduced when a macrophage is first challenged with UTP (A) and vice versa (B). C, summary of data. Note that there is no response to repeated UTP challenge when experiments are performed in Ca 2ϩ -free buffer, and there is less decrease of the second response when different G q -coupled receptor agonists are used. The number of single cells tested in each group is indicated in parentheses.
including cytoskeleton reorganization, gene expression, and the gating of various K ϩ channels (28 -31). Here we have identified the subtypes of P2Y and P2X receptors involved in transducing changes in extracellular nucleotide levels to transient Ca 2ϩ signaling in macrophages. We found that UTP-and ATP-induced Ca 2ϩ transients were abolished in macrophages isolated from P2Y 2 /P2Y 4 double knock-out or P2Y 2 Ϫ/Ϫ mice but were unaffected by genetic ablation of P2Y 4 . This conclusion was supported by RT-PCR analyses that showed that P2Y 2 , but not P2Y 4 , receptors were expressed in purified macrophages. Thus, in resident peritoneal macrophages, P2Y 2 is the only G protein-coupled receptor linking extracellular UTP and ATP to phospholipase C-␤, which catalyzes the hydrolysis of phosphatidylinositol 4,5-bisphosphate, releasing the second messengers IP 3 and diacylglycerol, an activator of PKC. Transient elevation of IP 3 promotes the release of Ca 2ϩ from the endoplasmic reticulum by increasing the open probability of IP 3 receptors (32). In addition, cytosolic [Ca 2ϩ ] regulates the various IP 3 receptor subtypes in a bell-shaped fashion (inhibition at higher concentrations), a gating property that may facilitate the oscillatory Ca 2ϩ release pattern we observed at submaximal P2Y receptor agonist concentrations.
In P2Y 2 -deficient macrophages, ATP-induced Ca 2ϩ mobilization was abolished in strictly Ca 2ϩ -free conditions. This lack of responsiveness to ATP underscores that there is no equivalent to the human P2Y 11 receptor in mouse macrophages. In the presence of extracellular Ca 2ϩ , a small Ca 2ϩ transient induced by ATP is observed and most likely represents the pure P2X receptor response to this agonist. We found that mouse macrophages express P2X 1 , P2X 4 , and P2X 7 mRNA, which seems to be a common pattern of expression for macrophages from various species (23,29,33). P2X 7 receptors do not form heteromeric receptors with other P2X subunits, but heteromeric P2X 1 /P2X 4 receptor assembly has been described recently in an oocyte expression system (34). Whether heteromeric P2X 1 / P2X 4 receptors, or even P2Y receptor heteromerization (35), play a functional role in native cells remains to be established. At this stage, we assume that homomeric P2X 4 receptors are the major contributors to Ca 2ϩ influx in the concentration range we tested (0.1-250 M) because the P2X 1 receptor response has been reported to terminate within 1 s, and the P2X 7 receptor is much less sensitive to ATP (15). Under physiological conditions, P2X receptor signaling is not only switched off by receptor desensitization, which we could observe in superfused cells, but also by rapid degradation of ATP by ecto-nucleotidases.
ADP is the preferential agonist for the G q -coupled P2Y 1 receptor, and RT-PCR analyses showed that this P2Y receptor subtype is expressed in mouse macrophages along with P2Y 2 . In peritoneal macrophages isolated from BALB/c mice, this agonist has been reported to have no effect on [Ca 2ϩ ] i at 10 M but to have a modest effect (possibly overestimated because of contaminating ATP) at 100 M (23). We also observed that ADP (purified with hexokinase) had negligible effect in macrophages isolated from WT or P2Y 2 -deficient mice. We additionally detected P2Y 6 receptors in RT-PCR analyses; however, UDPinduced Ca 2ϩ release was scant in resident macrophages isolated from either P2Y 2 /P2Y 4 double knock-out or WT mice. A modest response to UDP, though, was observed after macrophages had been activated with lipopolysaccharide and interferon-␥, suggesting that, functionally, P2Y 6 receptors are weakly expressed in resting macrophages but, following activation by stimuli such as Toll-like receptor ligands, P2Y 6 gene expression is switched on, increasing the scope of uridine nucleotide-induced signaling.
Signal transduction by G protein-coupled receptors, including P2Y receptors, is tightly controlled by mechanisms that dampen signal transmission. The rapid termination of P2Y 2 receptor-induced Ca 2ϩ signaling and refractoriness to repeated agonist challenge we observed could, in principle, be due to several mechanisms. The activated G q/11 subunit may be inhibited by GTPase-activating proteins, such as members of the regulators of G protein signaling family (17,36), and direct receptor phosphorylation by PKC or GRKs (promoting the recruitment of ␤-arrestin) may uncouple the receptor. Our data obtained using GRK2 ϩ/Ϫ and GRK6 Ϫ/Ϫ mice, as well as pharmacological inhibition of PKC, suggest that GRK6, PKC, or normal levels of GRK2 are not essential for switching off Ca 2ϩ signaling induced by P2Y 2 receptor activation. Consistent with our observations, González and co-workers (37, 38) deduced  NOVEMBER 17, 2006 • VOLUME 281 • NUMBER 46 that PKC was not responsible for agonist-induced P2Y 2 receptor desensitization, but they provided evidence that receptor phosphorylation by an unidentified kinase was involved. We also found that inhibition of CaMKII did not prevent apparent P2Y 2 receptor desensitization. Similarly, Tulapurkar et al. (39) reported that inhibition of CaMKII did not affect desensitization of P2Y 1 receptors, but it blocked internalization of activated receptors.

P2Y and P2X Receptors in Macrophages
Local depletion of phosphatidylinositol 4,5-bisphosphate (the source of IP 3 ), which has been nicely shown in myocytes to have low mobility in the surface membrane (40), could contribute to switching off signal transmission by activated G q -coupled receptors. This could explain why the Ca 2ϩ response is greater when the second agonist is C5a instead of UTP after initial challenge with UTP, i.e. the C5a receptor has access to a spatially distinct lipid pool. Downstream of IP 3 generation, ER Ca 2ϩ stores may become depleted, intraluminal Ca 2ϩ cycling may be rate-limiting, or the IP 3 receptor may be rendered rectractory. ER [Ca 2ϩ ] measurements suggest that Ca 2ϩ stores are rapidly replenished following P2Y 2 receptor stimulation, but under certain conditions favoring plasma membrane Ca 2ϩ transport systems (Ca 2ϩ -free buffer) ER stores are depleted, and there is no response to repeated agonist challenge. Thus, the interplay of Ca 2ϩ transport systems is an important determinant of G q -coupled receptor-mediated Ca 2ϩ signaling.
In conclusion, we can summarize the main findings of this study in the schematic diagram shown in Fig. 10. Various stimuli such as mechanical stress, cell injury, or inflammation release UTP and ATP into the extracellular space. Macrophages sense the increased nucleotide levels through the dominant G q -coupled P2Y 2 receptor that is activated by UTP ϭ ATP but insensitive to UDP and ADP. ATP additionally induces Ca 2ϩ influx via P2X receptors, and UDP activates the P2Y 6 receptor, functionally expressed in activated macrophages. Downstream of the activated G q -coupled receptors, IP 3 is generated and releases Ca 2ϩ from the endoplasmic reticulum. At the same time, the recently identified Ca 2ϩ -release sensor STIM1 probably translocates to the surface membrane (Fig.  10, dashed arrow) and activates Ca 2ϩ release-activated Ca 2ϩ channels to promote Ca 2ϩ entry and refilling of stores (25). Two Ca 2ϩ -ATPases compete to clear Ca 2ϩ from the cytosol as follows: the sarco(endo)plasmic reticulum Ca 2ϩ -ATPase and the plasma membrane Ca 2ϩ -ATPase. Conditions favoring plasma membrane Ca 2ϩ -ATPase activity will lead to diminished ER refilling.