Resensitization of P2Y Receptors by Growth Factor–Mediated Activation of the Phosphatidylinositol-3 Kinase in Retinal Glial Cells

Abstract

purpose. To determine whether activation of receptor tyrosine kinases enhances the responsiveness of purinergic P2Y receptors in Müller glial cells, known to induce Müller cell proliferation.

methods. The P2Y receptors of primary cultured Müller cells of the guinea pig were desensitized by short (30 seconds to 10 minutes)- and long (24 or 48 hours)-term application of adenosine 5′-triphosphate (ATP). Recordings of ATP-evoked intracellular calcium responses showed whether short-term application of different growth factors evoke a resensitization of the receptors. Coapplication of pharmacologic inhibitors showed whether activation of protein kinases is involved in receptor resensitization.

results. Both short- and long-term incubation with ATP induced a significant P2Y receptor desensitization, which was indicated by a strongly reduced intracellular calcium mobilization and which lasted for at least 48 hours. However, the receptors were significantly resensitized after short-term application of platelet-derived, epidermal, or nerve growth factors. The growth factor–mediated resensitization was dependent on an intact cytoskeleton and on the activation of protein phosphatases and of the phosphatidylinositol-3 kinase (PI3K), but was independent of the activation of protein kinase C, src kinases, or extracellular signal-regulated kinases.

conclusions. The results show that activation of receptor tyrosine kinases causes, via activation of PI3K and protein phosphatases, a resensitization of P2Y receptors formerly desensitized by agonist application. The growth factor–mediated resensitization may underlie the previously observed enhanced responsiveness of P2Y receptors in retinal glial cells in experimental retinal detachment and proliferative vitreoretinopathy and may contribute to the induction of reactive gliosis and Müller cell proliferation.

Detachment of the neural retina from the pigment epithelium induces complex responses of neuronal, glial, and pigment epithelial cells. In addition to photoreceptor cell degeneration and remodeling of retinal neurons, reactive gliosis and its consequences may contribute to the observed suboptimal recovery of visual acuity after reattachment surgery. ^{1 2} After experimental detachment, Müller cells begin to undergo hypertrophy, filling the space left by dying photoreceptor cells. They then grow into the subretinal space where they form glial scars. Müller cells also display a transient proliferative response that begins within hours and peaks at 3 to 4 days of detachment. ^{3 4 5} Among the secondary complications caused by Müller cell gliosis, are the development of subretinal fibrosis (which inhibits the regeneration of outer photoreceptor segments ^{2 6} ) and proliferative vitreoretinopathy (PVR), which may impair visual recovery. ^{1 2} Furthermore, a failure of glutamate recycling by the gliotic Müller cells has been reported. ⁷  

It has been shown that Müller cell gliosis is accompanied by distinct alterations of membrane currents ^{8 9 10 11} and receptor responsiveness, ^{11 12 13} both of which have been implicated in Müller cell proliferation. ^{14 15} In particular, during experimental detachment and PVR, the Müller cells display an enhanced responsiveness (intracellular calcium increases and calcium-dependent K⁺ currents) to extracellular adenosine 5′-triphosphate (ATP) ^{11 12 13} which fits with the observation that ATP stimulates the proliferation of cultured Müller cells, by acting on purinergic P2 receptors. ^{12 16 17} However, it remains to be resolved which mechanism(s) cause the enhanced Müller cell responsiveness on activation of P2 receptors in detached and PVR retinas. In the retinas of rabbits and guinea pigs, both ATP-evoked cytosolic calcium responses and ATP-stimulated proliferation of Müller cells are mediated by the activation of G-protein-coupled receptors of the P2Y family. ^{11 16} G-protein-coupled receptors are known to desensitize rapidly on agonist application. Thus, an inhibition of receptor desensitization may represent one mechanism of enhanced responsiveness. The present study was performed to determine whether altered receptor desensitization contributes to the enhanced responsiveness to ATP in gliotic Müller cells. Because growth factors have been crucially implicated in gliotic responses and proliferation during retinal detachment and PVR ^{5 18 19} and because P2Y receptor stimulation in cultured Müller cells results in autocrine/paracrine release of growth factors, ¹⁷ we investigated whether activation of receptor tyrosine kinases by growth factors causes receptor resensitization of P2Y receptors desensitized by ATP. 

Materials and Methods

Platelet-derived growth factor (PDGF-AB/BB), epidermal growth factor (EGF), and nerve growth factor (NGF), PD98059, LY294002, AG1478, and Gö6976 were obtained from Calbiochem (Bad Soden, Germany). Nagarse (subtilisin, EC 3.4.21.14) and ATP (sodium salt) were from Serva (Heidelberg, Germany); fura-2/acetoxymethylester (AM) was from Molecular Probes (Eugene, OR), and fetal calf serum was from Biochrom (Berlin, Germany). All other substances were obtained from Sigma-Aldrich (Deisenhofen, Germany). 

Cell Culture

Animal care and handling were performed in accordance with applicable German laws and with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. A table of the drugs used in this study and their targets. Primary cultures of Müller glial cells were obtained from guinea pigs (250–400 g), as described previously. ¹⁶ The animals were deeply anesthetized (urethane; 2.0 g/kg, intraperitoneally) and decapitated, the eyes were enucleated, and the excised retinas were incubated in phosphate buffer containing nagarse (1 mg/mL) for 30 minutes at 37°C. After a wash and treatment with DNase I (200 U/mL), the retinas were dissociated mechanically, and the isolated cells were plated on uncoated coverslips (diameter 15 mm; Glaswarenfabrik Hecht, Sontheim/Rhön, Germany). The cells were cultured in minimum essential medium (MEM) supplemented with 10% serum at 37°C in 5% CO₂. After 8 days in culture, test substances were added to serum-free MEM 24 or 48 hours before the cells were used for calcium imaging experiments. Lipophilic substances were dissolved in dimethylsulfoxide, which was shown to have no effect on the calcium responses. Approximately 96% of the cells displayed immunoreactivity for vimentin, and approximately 99% were positively immunolabeled for glial fibrillary acidic protein. ²⁰ Because the guinea pig retina lacks astrocytes, most of the cells were considered to represent Müller glial cells. 

Calcium Imaging

For fluorescence measurements, the cells were loaded with 2 μM fura-2/AM for 30 minutes at 37°C. Measurements were performed at room temperature, by using a bath solution that contained (in mM) 129 NaCl, 3 KCl, 1 CaCl₂, 0.2 MgSO₄, 20 glucose, and 10 HEPES (pH 7.4, adjusted with NaOH). Before application of test substances, the cells were perfused for 20 minutes with the bath solution. Fluorescence was excited at 340 (F ₃₄₀) and 380 nm (F ₃₈₀) and recorded above 510 nm. Images were taken every 6 seconds and analyzed by with an imaging system (Till-Photonics; Munich, Germany; consisting of Polychrom IV, SensiCam, and Till Vision Software). 

Data Presentation

Data were analyzed on computer (SigmaPlot 2000; SPSS Inc., Chicago, IL). The fluorescence ratio R (F ₃₄₀/F ₃₈₀) was calculated to describe the relative level of the cytosolic free calcium concentration; an increase in R indicates an increase in free calcium. The ΔR values represent the amplitudes of the agonist-evoked calcium responses and were calculated by subtraction of the mean ratio—measured within 1 minute before the application of test substances—from the maximum ratio measured during application and the following 2 minutes. Significant differences were evaluated by Mann-Whitney test. Statistical significance was accepted at P < 0.05. 

Results

P2Y Receptor-Evoked Calcium Responses

Stimulation of P2Y receptors in primarily cultured Müller cells of the guinea pig by extracellular application of ATP resulted in a transient increase in cytosolic free calcium, which was dependent on the activation of phospholipase C and on the release of calcium ions from inositol 1′,4′,5′-triphosphate (IP₃)-gated intracellular stores. The intracellular calcium release secondarily evoked a transient calcium influx from the extracellular space. ¹⁶ In addition to ATP (EC₅₀ 0.6 μM), uridine 5′-triphosphate (UTP), adenosine 5′-diphosphate (ADP), and uridine 5′-diphosphate (UDP) were found to evoke intracellular calcium responses in a dose-dependent manner, with mean EC₅₀ of 0.6 μM (UTP), 0.5 μM (ADP), and 66 μM (UDP) (Fig. 1A) . Application of 2′-3′-O-(4-benzoylbenzoyl)-ATP (50 μM), a more selective agonist of P2X receptors, did not evoke calcium responses in the cells tested (not shown). Preincubation of the putative nonselective P2 receptor antagonist pyridoxal phosphate 6-azophenyl-2′,4′-disulfonic acid (PPADS; 200 μM) for 5 minutes depressed the ATP (50 μM)-evoked calcium transients almost entirely, whereas suramin (200 μM) had no effect (not shown). 

P2Y Receptor Desensitization

G-protein-coupled receptors are rapidly desensitized by agonist exposure. To examine the desensitization characteristics of P2Y receptors in Müller cells, we performed short (30 seconds to several minutes)- and long (24 or 48 hours)-term incubation of cultured Müller cells with P2 receptor agonists and measured the amplitudes of the calcium responses evoked by ATP (500 μM). As illustrated in Figure 1B , application of ATP (500 μM) for 10 minutes resulted in an almost complete desensitization of the P2Y receptors, indicated by the strongly depressed calcium response during a second exposure to ATP. Similarly, the repetitive application of ATP (500 μM) for 30 seconds caused partial receptor desensitization as reflected by decreased amplitudes of the calcium responses (Fig. 1C) . Long-term exposure of ATP (500 μM) resulted in a dramatic receptor desensitization that lasted for at least 48 hours (Fig. 2A) . After a 24-hour exposure, the incidence of responding cells decreased by 89%, whereas the amplitude of the ATP-evoked increases in calcium decreased by 99% compared with control samples. The effect of long-term ATP exposure was concentration dependent, with a half-maximum effect at ∼20 μM (Fig. 2A) . A 24-hour exposure to ATP (500 μM) also desensitized the calcium responses evoked by UTP (200 and 500 μM) and ADP (500 μM) (data not shown). After a 48-hour exposure to ATP (500 μM), the incidence of responding cells increased (to 37% of the cells investigated), but the amplitude of the calcium response remained strongly depressed (by 92%; Fig. 2A ). 

Investigation of the putative mechanisms underlying the partial receptor desensitization evoked by repetitive application of ATP revealed that it was not mediated by activation of phosphatidylinositol-3 kinase (PI3K) or calcium-dependent isoforms of protein kinase C (PKC), as indicated by the lack of effects of the selective inhibitors LY294002 and Gö6976, respectively (Fig. 1C) . Similarly, the broad-spectrum protein kinase inhibitor H-7 (25 μM) was without effect (not shown). Likewise, the receptors remained partially desensitized after activation of PKC by phorbol 12-myristate 13-acetate (PMA; Fig. 1C ). 

To determine whether P2Y receptor desensitization is due to the depletion of calcium from intracellular stores, cells were cultured for 24 hours in medium containing ATP (500 μM), and the basal cytosolic calcium level and the amount of calcium present in the intracellular stores were determined by the method explained in the legend of Figure 2B . Exposure to ATP changed neither the amount of calcium present in the cytosol under unstimulated conditions (Fig. 2C) , nor the calcium content of the intracellular stores (Fig. 2D) . However, addition of serum (10%) to the medium induced a significant elevation of the cytosolic calcium concentration (by ∼15%, P < 0.001) and partially depleted the intracellular calcium stores (by ∼25%; P < 0.001). 

As several different P2 receptor agonists were shown to elicit intracellular calcium responses (Fig. 1A) , we considered whether P1 and P2 receptor agonists may also cause purinergic receptor desensitization. This was investigated by recording the ATP-evoked calcium responses after incubation of the cells for 24 hours in the presence of the respective agonists. As shown in Figure 3B , adenosine, 2-methylthio-ATP (MeSATP), ADP, and UTP exerted no effect on the amplitude of the ATP-evoked calcium response (Fig. 3A) . Coculturing of the cells with ATP and Gö6976, an inhibitor selective for the PKCα, -β, and -γ isoforms did not affect the desensitizing effect of ATP. 

To determine whether exposure to ATP or P2 receptor agonists evokes heterologous desensitization among different subtypes of G-protein-coupled receptors, we investigated the calcium transients evoked by sphingosine 1-phosphate (S1P). S1P acts through the activation of G-protein-coupled receptors encoded by genes belonging to the Edg (endothelial differentiation gene) family. ²¹ Application of S1P (0.1 μM) to the cultured Müller cells evoked an intracellular calcium response that was completely blocked in the presence of the phospholipase C inhibitor U73122 (Fig. 3B) , suggesting that, like the P2Y receptors, activation of the Edg receptors results in an evoked calcium release from IP₃-gated stores of the Müller cells. Culturing the cells for 24 hours in the presence of ATP (20 or 500 μM) did not result in any alteration of the S1P-evoked calcium transients, suggesting heterologous receptor desensitization (Fig. 3C) . With the exception of adenosine, which displayed a slight depressive effect, neither of the P2 agonists tested caused a desensitization of the calcium responses evoked by S1P (Fig. 3C) . 

Growth Factor–Evoked P2Y Receptor Resensitization

Growth factors have been shown to resensitize P2Y receptors in a range of different cell systems, ²² via phosphorylation by PKC. ^{23 24} Because growth factors have been crucially implicated in the retinal responses during detachment and PVR, ^{5 18 19} we investigated the effects of EGF, PDGF, and NGF on the ATP-induced desensitization of P2Y receptors in Müller cells. EGF (in a concentration range between 50 and 200 ng/mL) evoked transient calcium responses in 50% to 70% of the cells investigated, with a maximum effect at 100 ng/mL (Figs. 4B 4E) . PDGF (50 or 100 ng/mL) elicited only a slight elevation in the intracellular free calcium (Fig. 4C)in approximately 10% of the cells investigated (Fig. 4E) . However, both EGF and PDGF caused significant elevations of the ATP (500 μM)-evoked calcium response when they were applied for 10 minutes before ATP application in cells in which the P2Y receptors had been desensitized by culturing for 24 hours in the presence of ATP (500 μM). In the absence of growth factor, both the incidence and the amplitude of the ATP-evoked calcium responses were small, because of the desensitization of the P2Y receptors (Fig. 4A) . When the cells were exposed to PDGF (50 or 100 ng/mL) or EGF (100 or 200 ng/mL) before application of ATP, the incidence (Fig. 4E)and the amplitude of the ATP-evoked calcium response were significantly enhanced (Figs. 4B 4C) . A similar resensitizing effect of PDGF was observed during repeated short-term application of ATP (not shown). NGF (100 ng/mL) was observed to cause a resensitization of P2Y receptors in a manner similar to PDGF, as evidenced by the reappearance of a small-amplitude calcium response in ∼41% of the cells investigated (Figs. 4D 4E) . 

Mechanism of Growth Factor–Evoked Resensitization

In the case of EGF, a resensitizing effect was observed at the higher concentrations tested (100 and 200 ng/mL) but not at 50 ng/mL, although the latter concentration was sufficient to elicit a calcium response (Fig. 4E) . Accordingly, PDGF also evoked a receptor resensitization in most cases in which no apparent calcium response was triggered (Fig. 4C) . Together, these observations suggest that the growth factor–evoked resensitization is independent of any growth factor–evoked calcium response. To elucidate the intracellular signaling pathways mediating the resensitizing effect of EGF, we tested a series of selective pathway blockers. The selective inhibitor of the EGF receptor tyrosine kinase, tyrphostin AG1478, ²⁵ completely blocked both the EGF-induced calcium response and the P2Y receptor resensitization (Figs. 5A 5D)whereas the PDGF-induced resensitization remained unaffected by AG1478 (not shown), suggesting that the activation of the specific receptor tyrosine kinase is a necessary step in the signaling chain of the EGF effect. Concanavalin A, an inhibitor of receptor internalization, ^{26 27} had no effect on either the amplitude of the EGF-evoked intracellular calcium transient or the EGF-evoked resensitization (Fig. 5D) . Monensin, a fungal sodium ionophore that has been shown to inhibit EGF receptor recycling from endosomes, ²⁸ was observed to cause transient increases of intracellular free calcium but was without effect on the EGF-evoked P2Y receptor resensitization (Fig. 5D) . The lack of effect of monensin and concanavalin A suggests that receptor internalization and subsequent recycling, induced by binding of EGF, were not involved in the P2Y receptor resensitization. However, an inhibitor of microtubular polymerization, nocodazole (which also induced a transient calcium response) significantly decreased the EGF-evoked P2Y receptor resensitization, with decreases in both the percentage of cells displaying resensitization and the amplitude of the ATP-evoked calcium responses (not shown). Nocodazole had no effect on the ATP-evoked calcium transients in control cells (not shown). It is concluded that the EGF-evoked resensitization is dependent on an activation of the EGF receptor tyrosine kinase, and requires an intact cytoskeleton. 

Several different blockers of protein kinases failed to inhibit the resensitizing effect of EGF when cultured Müller cells were given a 2-minute incubation with these compounds, and they were coapplied with the growth factor H-7 (a broad-spectrum protein kinase inhibitor), Gö6976 (an inhibitor of the calcium-dependent PKC isoforms), PD98059 (an inhibitor of the mitogen-activated protein kinase kinase [MEK1]), and PP2 (an inhibitor of src family tyrosine kinases; Fig. 5D ). However, in the presence of the PI3K inhibitor LY294002, the EGF-mediated resensitization of ATP-evoked calcium responses was blocked (Figs. 5B 5D) . As the phosphorylation state of G protein-coupled receptors is known to be regulated by protein phosphatases, we tested the effects of two phosphatase inhibitors, cyclosporine (an inhibitor of phosphatase type 2B) and okadaic acid (an inhibitor of phosphatases type 1 and 2A) on the EGF-induced P2Y receptor resensitization. Preincubation with cyclosporine for 2 minutes and its coapplication with the growth factor did not result in any alteration of the resensitizing effect of EGF (not shown); however, okadaic acid exerted a significant (P < 0.05) depression (Fig. 5C) . We conclude that the EGF-evoked receptor resensitization is mediated by activation of PI3K and of protein phosphatases. 

Discussion

Primarily cultures of Müller cells from the guinea pig retina apparently express several different subtypes of metabotropic P2Y receptors, whereas there is presently no indication of ionotropic P2X receptor expression. ¹⁶ The calcium imaging experiments presented herein suggest that, among others, the P2Y₁, P2Y₂, and P2Y₄ receptors are expressed by the cells, and MeSATP, ATP and UTP, and UTP, respectively, are their selective agonists. Exposure to ATP resulted in a fast and sustained (at least, up to 48 hours) desensitization of ATP- and UTP-evoked intracellular calcium signaling. This desensitization was reflected in the decrease of both the incidence of responding cells and the amplitude of the calcium transients and was apparently independent of the activation of PKC or PI3K. The ATP-induced receptor desensitization is not caused by intracellular calcium depletion and fails to affect the sensitivity to agonists for other calcium-mobilizing receptors such as the Edg receptors, which are activated by S1P. In other cell systems, an agonist-induced desensitization of P2Y receptors has been demonstrated, which was mediated by multiple intracellular pathways, depending on the receptor subtype investigated, either with or without involvement of PKC or PKA activation. ^{29 30 31 32} In Müller cells, the intracellular signaling pathways that underlie agonist-induced receptor desensitization remain to be established. Moreover, the subtypes of P2 receptor that mediate this response remain to be determined. Exposure to MeSATP and UTP did not desensitize the receptors, suggesting that the desensitizing effect of ATP was not mediated by any of the identified subtypes (P2Y₁, P2Y₂, and P2Y₄). However, the possibility of an agonist-specific desensitization cannot be ruled out, which means that different agonists may act at the same receptor subtype but may cause its desensitization with different efficacy. 

Previously, we demonstrated that there is cross talk between the signaling cascade of P2Y receptors and receptor tyrosine kinases that results in the stimulation of Müller cell proliferation. ¹⁷ This cross talk involves the release of growth factors (PDGF and heparin-binding EGF-like growth factor, an endogenous agonist of the EGF receptor) and subsequent transactivation of the respective receptor tyrosine kinases. In the present study, there was also cross talk from the signaling cascades of receptor tyrosine kinases toward the P2Y receptors in Müller cells. Activation of receptor tyrosine kinases reversed the agonist-induced desensitization of P2Y receptors, resulting in both an increased incidence of responding cells and enhanced amplitude of the ATP-evoked calcium transients. This prompt heterologous receptor resensitization suggests that transcriptional downregulation is not the only cause of the ATP-induced long-term desensitization of P2Y receptors. 

Desensitized receptors may be functionally recovered by different mechanisms. One mechanism is de novo synthesis and expression of receptor protein. It has been shown that, after internalization of P2Y₂ receptors, protein synthesis is necessary for full recovery of the amount of receptors expressed at the cell surface. ³³ However, de novo protein synthesis takes ∼24 hours. ³³ Similarly, the mitogen-activated protein-kinase–dependent upregulation of P2Y₂ receptors in vascular smooth muscle cells has been shown to take several hours, after stimulation with growth factors. ³⁴ Because in the present experiments, resensitization of the ATP-evoked calcium responses was observed within minutes after growth factor application, we assume that de novo synthesis of receptor protein is not the main mechanism of the growth factor effects on P2Y receptors in Müller cells. 

The resensitizing effect of EGF was found to be independent of the growth factor–evoked calcium response as well as the activation of PKC, the src family tyrosine kinases, and MEK. However, in the present study, this effect was (1) dependent on tyrosine phosphorylation of the growth factor receptor, (2) required an intact cytoskeleton, and (3) involved activation of the PI3K and protein dephosphorylation as necessary steps. To our knowledge, this is the first description of a cellular system where activation of PI3K by receptor tyrosine kinases causes an enhanced responsiveness of P2Y receptors. In other cell systems, activation of PKC has been found to be essential for growth factor–induced P2Y receptor resensitization. ^{23 24} Moreover, a blocking effect of okadaic acid on P2Y₂ receptor resensitization has been observed in several different cell systems. ^{32 35} The activation of both PI3K and protein phosphatases may result in alterations of the functional state of the receptors or of the functional coupling of the receptors to the calcium signaling pathway. Although this question remains to be resolved, we conclude that the desensitization of P2Y receptors in Müller cells is under the control of growth factors. Thus, our findings support the hypothesis that growth factors potentiate the action of nucleotides as mediators of Müller cell gliosis and proliferation. 

Both PKC and PI3K have been implicated in the mediation of mitogenic signals after P2Y receptor stimulation in Müller cells. ^{16 17} The mitogenic effect of ATP on Müller cells was mediated by activation of ERKs via the Ras-Raf-MEK pathway and was significantly suppressed (by ∼60%) when the activation of the PI3K was pharmacologically inhibited. ¹⁷ The functional significance of the growth factor–evoked increase of ATP-induced calcium signaling is underlined by the previous observation that the proliferation rate of cultured Müller cells correlates positively with the duration of the ATP-evoked calcium transients. ¹⁶  

During experimental retinal detachment and PVR in the rabbit, Müller cells displayed enhanced responsiveness after activation of P2Y receptors. In particular, the incidence of cells responding with an increase of the cytosolic free calcium after extracellular application of ATP increased significantly in a manner that appears to be dependent on the duration of detachment, but even more so, during PVR. ^{11 15} The mechanisms underlying the increased responsiveness of Müller cells to ATP during detachment and PVR are unknown. A transcriptional regulation, resulting in enhanced receptor protein expression, or other mechanisms may be involved. The results of the present study suggest that a growth factor–induced increase in receptor resensitization may contribute to the enhanced responsiveness of Müller cells to ATP. The responsiveness of Müller cells to ATP decreased on P2-receptor activation, which thereby caused a termination of the mitogenic effect. Furthermore, ATP was degraded by ectonucleases. Both mechanisms should restrict the effects of ATP in both a spatial and temporal manner when ATP is released during detachment or other pathologic conditions. However, growth factors (and other signaling molecules that may also provoke a resensitization of P2Y receptors) may abolish these limitations because they are less rapidly inactivated. Growth factors additionally enhanced proliferation after further pathologic events involving ATP release, even at concentrations below the level where they induced proliferation by themselves. 

Basic fibroblast growth factor (FGF-2) is released into the retina during the first minutes of experimental detachment ⁵ and, among other growth factors, is thought to contribute to the development of PVR. ^{18 19} The signaling cascade evoked by PDGF also plays a major role in experimental PVR. ^{36 37} It has been shown that the PDGF-evoked activation of PI3K, and to a lesser extent phospholipase C-γ, are necessary for initiation of PVR, whereas the activation of the src family tyrosine kinases is not necessary for its development. ³⁸ We have shown in this study that the resensitizing effect of EGF was mediated by PI3K but was independent of the src family of tyrosine kinases. Although further detailed (in vivo) research remains to be undertaken, this congruence supports the suggestion that the growth factor–PI3K signaling cascade contributes to the development of PVR (perhaps, among other pathways) by enhancing the responsiveness of other types of pathogenic receptors to their agonists, generating a signaling avalanche that eventually results in exacerbated mitogenic stimulation. 

 Supported by grants from the Bundesministerium für Bildung und Forschung (IZKF at the University of Leipzig, 01KS9504, Projects C5 and C21), and by Grant Re849/8-3 from the Deutsche Forschungsgemeinschaft.

 Submitted for publication April 14, 2004; revised August 12, 2004; accepted August 17, 2004.

 Disclosure: M. Weick, None; P. Wiedemann, None; A. Reichenbach, None; A. Bringmann, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Short-term application of ATP (500 μM) induced P2Y receptor desensitization in Müller cells. (A) Dose–response relationships of different P2 receptor agonists. The amplitudes of the agonist-induced calcium transients (ΔR) were measured. (B) Exposure to ATP for 10 minutes caused a nearly complete absence of the calcium response on a second agonist application. (C) The receptor desensitization was not mediated by activation of PI3K or PKC, as indicated by the absence of effects of the selective blockers LY294002 (25 μM) and Gö6976 (200 nM), respectively. Likewise, the PKC activator PMA (1 μM) had no effect. LY294002 and Gö6976 evoked slight increases of the basal cytosolic calcium concentration but had no effects on the agonist-evoked calcium transients. Mean (±SD) curves for 30 to 85 cells.

Figure 2.

 

Long-term exposure to ATP caused desensitization but did not change the calcium content of Müller cells. (A) Mean (±SEM) incidence of cells showing calcium responses on extracellular application of ATP (500 μM; left) and mean (±SEM) amplitude of the ATP-evoked calcium transients (ΔR; right). The cells were cultured for 24 or 48 hours in the absence or presence of ATP at the different concentrations shown. Inset: Mean (±SD) records of cells examined after 24 hours of culturing in the absence and presence of ATP (20 and 500 μM, respectively). The data were derived from 49 to 225 cells in 3 to 10 independent experiments. *P < 0.05; ***P < 0.001, versus control. (B–C) Cells were cultured for 24 hours in the absence or presence of ATP (500 μM) or fetal calf serum (10%). Thereafter, the cytosolic calcium concentration and the amount of calcium present in the intracellular stores were determined by fura-2 imaging. (B) Example of a record in one cell. The relative calcium concentration in the cytosol was estimated by noting the fluorescence ratio F ₃₄₀/F ₃₈₀ at the beginning of the record (R ₀). The relative amount of calcium present in the intracellular stores was estimated by integration of the calcium transient, which was induced by cyclopiazonic acid (4 μM) in nominally calcium-free extracellular solution (∫_t R). Cyclopiazonic acid is an inhibitor of ER calcium-ATPase, thus causing a slow depletion of intracellular stores from calcium. (C) Mean (±SEM) relative cytosolic calcium concentration. (D) Mean (±SEM) relative amount of calcium present in the stores. The values were obtained from between 526 and 943 cells. *P < 0.001 versus control.

Figure 3.

 

Long-term exposure to ATP caused desensitization of P2Y receptors in Müller cells, but has no effect on the functional state of receptors for sphingosine 1-phosphate (S1P). (A) Effect of 24-hour culturing in the absence and presence of different P2 receptor agonists (20 μM) on the amplitude of the intracellular calcium transients evoked by ATP (500 μM). The PKC inhibitor Gö6976 (200 nM) did not reverse the effect of ATP (20 μM) on the receptor desensitization when it was present in the culture medium for 24 hours. (B) Application of ATP (500 μM) and S1P (0.1 μM) evoked transient intracellular calcium responses. In the presence of the phospholipase C inhibitor U73122 (4 μM), the S1P-evoked calcium transient was largely depressed. Mean (±SD) records from 36 control cells. (C) Effect of a 24-hour culturing period in the absence and presence of different P2 receptor agonists on the amplitude of the intracellular calcium transients evoked by S1P (0.1 μM). The data were derived from each 49 to 225 cells in 3 to 10 independent experiments. *P < 0.05, ***P < 0.001 versus control.

Figure 4.

 

Activation of receptor tyrosine kinases resulted in increased ATP (500 μM)-evoked calcium responses in Müller cells. The calcium responses were recorded after a 24-hour incubation in the presence of ATP (500 μM) which resulted in the desensitization of the P2Y receptors. (A) Control record displaying calcium responses of small amplitude on application of ATP. (B) Preapplication of EGF (100 ng/mL) resulted in a significantly increased ATP-evoked calcium response. (C, D) Similar effects were observed after application of PDGF (100 ng/mL) or NGF (50 ng/mL). (E) Mean (±SEM) incidence of cells that responded to ATP (500 μM) and to PDGF (100 ng/mL), NGF (100 ng/mL), or EGF (50, 100, and 200 ng/mL), with a transient elevation of cytosolic calcium. ATP was applied before (1. ATP) and after a 10-minute incubation with the growth factors (2. ATP). The data were obtained from 3 to 11 independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 5.

 

The EGF-evoked resensitization of P2Y receptors was mediated by activation of the EGF receptor tyrosine kinase, of PI3K, and protein phosphatases. The cells were cultured for 24 hours in the presence of ATP (500 μM) before calcium imaging experiments. (A) The tyrphostin AG1478 (300 nM) inhibited the EGF (100 ng/mL)-evoked calcium response as well as the growth factor–induced P2Y receptor resensitization. ATP was applied at 500 μM. (B) Coapplication of the PI3K inhibitor LY294002 (50 μM) and EGF (100 ng/mL) abolished the growth factor–induced resensitization of the ATP (500 μM)-evoked calcium response. (C) The phosphatase inhibitor okadaic acid (40 nM) inhibited the effect of EGF (100 ng/mL) on the ATP (500 μM)-evoked calcium response. (D) Mean (±SEM) amplitudes of the ATP (500 μM)-evoked calcium responses in Müller cells. The calcium responses were recorded in the absence and presence of a 10-minute preincubation of EGF (100 ng/mL) before ATP application. The following pharmacological agents were tested by coapplication with EGF: AG1478 (1 μM), LY294002 (50 μM), PP2 (100 nM), PD98059 (10 μM), Gö6976 (100 nM), H-7 (25 μM), monensin (100 μM), and concanavalin A (0.5 mg/mL). The data were obtained from 3 to 13 independent experiments. *P < 0.05.

Supplementary Materials

Supplementary Table S1 - 52.1 KB 

The authors thank John Wellard for a critical reading of the manuscript and Jana Krenzlin for excellent technical assistance.