ATP stimulates GRK-3 phosphorylation and beta-arrestin-2-dependent internalization of P2X7 receptor

Abstract

The objective of this study was to understand the mechanisms involved in P2X(7) receptor activation. Treatments with ATP or with the P2X(7) receptor-specific ligand 2',3'-O-(4-benzoylbenzoyl)adenosine 5'-triphosphate (BzATP) induced pore formation, but the effect was slower in CaSki cells expressing endogenous P2X(7) receptor than in human embryonic kidney (HEK)-293 cells expressing exogenous P2X(7) receptor (HEK-293-hP2X(7)-R). In both types of cells Western blots revealed expression of three forms of the receptor: the functional 85-kDa form present mainly in the membrane and 65- and 18-kDa forms expressed in both the plasma membrane and the cytosol. Treatments with ATP transiently decreased the 85-kDa form and increased the 18-kDa form in the membrane, suggesting internalization, degradation, and recycling of the receptor. In CaSki cells ATP stimulated phosphorylation of the 85-kDa form on tyrosine and serine residues. Phosphorylation on threonine residues increased with added ATP, and it increased ATP requirements for phosphorylation on tyrosine and serine residues, suggesting a dominant-negative effect. In both CaSki and in HEK-293-hP2X(7)-R cells ATP also increased binding of the 85-kDa form to G protein-coupled receptor kinase (GRK)-3, beta-arrestin-2, and dynamin, and it stimulated beta-arrestin-2 redistribution into submembranous regions of the cell. These results suggest a novel mechanism for P2X(7) receptor action, whereby activation involves a GRK-3-, beta-arrestin-2-, and dynamin-dependent internalization of the receptor into clathrin domains, followed in part by receptor degradation as well as receptor recycling into the plasma membrane.

Figures

Fig. 1

Effects of ATP (250 μM; A and B) and 2′,3′-O-(4-benzoylbenzoyl)adenosine 5′-triphosphate (BzATP, 100 μM; C and D) (arrows) on levels of cytosolic calcium (Cai, solid lines) and influx of ethidium bromide [Flu, dashed lines; in arbitrary units (AU)] in CaSki cells attached on filters for 6 days. B and D: means (±SD, 3–5 experiments in each point) of net increases in Cai (B) and ethidium bromide fluorescence (D) in response to ATP or BzATP. Extracellular calcium (Cao) was lowered to <0.1 mM by the addition of 1.2 mM EGTA. 1-[N,O-bis(5-isoquinolinesulfonyl)-N-methyl-l-tyrosyl]-4-phenylpiperazine (KN-62), oxidized ATP (oATP), and polyethylene glycol 600 (PEG-6000) were added at 100 nM, 75 μM, and 1 mM, respectively, 10–15 min before ATP was added. Change in Cai (ΔCai) above baseline for the late sustained increase in Cai was determined 30 min after adding ATP in cells pretreated with 75 μM suramin or pyridoxal phosphate-6-azophenyl-2′,4-disulfonic acid (PPADS) to block the transient acute increases in Cai (20, 21).

Fig. 2

Effects of ATP (A and B) and BzATP (C and D) (arrows) on levels of Cai (solid lines) and influx of ethidium bromide (dashed lines) in HEK-293 cells expressing exogenous human P2X7 receptor (HEK-293-hP2X7-R cells) attached to filters for 3 days. Experiments (repeated 3–5 times) were done as described in Fig. 1.

Fig. 3

A: reversibility of ATP effect. Six-day fura-2-loaded CaSki cells attached on filters were treated with 250 μM ATP, and at different time intervals the bathing solutions were replaced with fresh medium. Experiments were repeated 3–7 times, with similar trends. B: concentration-dependent effects (means of 2 experiments) of ATP and BzATP on Cai and ethidium bromide fluorescence in day 6 CaSki cells attached on filters. ΔCai values for the late sustained increase in Cai were determined 30 min after ATP was added in cells pretreated for 15 min with 75 μM suramin and PPADS. Dashed lines, changes in fluorescence 30 min after ethidium bromide was added.

Fig. 4

A: immunostaining of day 6 CaSki and HEK-293-hP2X7-R cells for the P2X7 receptor protein (×20). CaSki cells were treated 30 min before staining with 1 of the indicated concentrations of ATP. +Ag, coincubation with the P2X7 antigen. B: cellular distribution of the P2X7 receptor in day 6 CaSki cells as determined using confocal laser scanning microscopy (×40). a: Coincubation of anti-P2X7 receptor antibody with P2X7 antigen. b: Nuclear stain. c: Immunostaining with anti-P2X7 receptor antibody. d: Combined immunostaining of nuclei with anti-P2X7 receptor antibody. C: Western immunoblot analysis of P2X7 receptor protein using total homogenates from day 6 cultured CaSki (a) and HEK-293 (b) cells. In experiments with HEK-293 cells, we used wild-type cells, cells transfected with β-arrestin-2-GFP (β-Arr-2-GFP), or cells transfected with the full-length human P2X7 receptor (hP2X7-R) alone or in combination with β-arrestin-2-GFP. The experiments were repeated 2–5 times, with similar trends observed.

Fig. 5

Top: ATP effects on the distribution of P2X7 receptor isoforms in the plasma membrane and cytosol. Day 6 CaSki cells were treated with 250 μM ATP (arrows), and membrane-enriched and cytosolic fractions were prepared at time intervals of 0–30 min after treatment. Fifteen-microgram samples of protein were fractionated using gel electrophoresis and assayed using Western immunoblot analysis for the P2X7 receptor. The experiments were repeated twice, with similar trends observed. Middle and bottom: densitometric analysis of the data at top; for each isoform the density levels were normalized to the level of expression at time 0 before ATP was added.

Fig. 6

ATP effects (arrows) on the distribution of P2X7 receptor isoforms in the plasma membrane and cytosol of HEK-293-hP2X7-R cells. Top: immunoblot analysis. Middle and bottom: densitometric analysis of the data at top. The experiments were repeated twice, with similar trends observed, and data analysis was done as described in Fig. 5.

Fig. 7

ATP effects (arrows) on the distribution of P2X7 receptor isoforms in the plasma membrane and cytosol of HEK-293-c-Myc-hP2X7-R cells. Top: immunoblot analysis. Middle and bottom: densitometric analysis of the data at top. The experiments were repeated twice, with similar trends observed, and data analysis was done as described in Fig. 5, except that Western blot analysis was performed with anti-c-Myc antibody.

Fig. 8

ATP-induced phosphorylation of the P2X7 receptor. A, left: day 6 CaSki cells were labeled with [32P]orthophosphate and treated with 250 μM ATP for 5 min. Cell lysates were fractionated using gel electrophoresis and immunoprecipitated with the anti-P2X7 antibody. Right: Western immunoblots (IB) with anti-P2X7 antibody of parallel protein samples. B: day 6 CaSki cells were treated with 250 μM ATP; at time intervals of 0–30 min after treatment cells lysates were immunoprecipitated (IP) in a mixture of antibodies containing anti-phosphotyrosine PY20 and P99 antibodies and anti-phosphoserine and anti-phosphothreonine (PST) antibodies and immunoblotted with the anti-P2X7 antibody. C: day 6 CaSki cells were treated for 1 min with ATP at concentrations ranging from 0 to 500 μM. Cells lysates were immunoprecipitated with the indicated anti-phosphorylation antibodies, alone or in combination, and immunoblotted with the anti-P2X7 antibody. The experiments were repeated twice, with similar trends observed.

Fig. 9

ATP-induced colocalization of the P2X7 receptor with GRK-3 (A), β-arrestin-2 (B), and dynamin and clathrin (C). Day 6 CaSki cells (A–C) or HEK-293-c-Myc-hP2X7-R cells (A and B) were treated with 250 μM ATP, and cell lysates were immunoprecipitated or immunoblotted at time intervals of 0–30 min after treatment as indicated. D: effects of treatment with ATP on total cellular levels of the P2X7 receptor forms in day 6 CaSki cells. Each experiment was repeated 2–3 times, with similar trends observed.

Fig. 10

A: ATP-induced recruitment of β-arrestin-2 into submembranous regions: CaSki cells were transfected with β-arrestin-2-GFP and analyzed using real-time confocal microscopy. a: Phase images. b: Nuclei staining. c–e: Fluorescence images at excitation wavelength of 488 nm (×20) at baseline (c) and 3 (d) and 5 (e) min after 250 μM ATP was added. The experiment was repeated 3 times, with similar trends observed. B: a and b: low magnification (×20) of HEK-293-hP2X7-R cells cotransfected with β-arrestin-2-GFP (a, nuclei staining; b, fluorescence at excitation wavelength of 488 nm). c–e: Higher magnification (×40). c: HEK-293 cells transfected with β-arrestin-2-GFP only. d: HEK-293 cells transfected with β-arrestin-2-GFP after treatment with 250 μM ATP. e: HEK-293-hP2X7-R cells transfected with β-arrestin-2-GFP. C: ATP-induced recruitment of β-arrestin-2 into submembranous regions: HEK-293-hP2X7-R cells cotransfected with β-arrestin-2-GFP and treated with 10 μM angiotensin (a–c), 10 μM UTP (d–f), or 250 μM ATP (g–i). Real-time confocal laser microscopy was used to determine fluorescence at excitation wavelength of 488 nm at baseline (a, d, g) or 5 (b, e, h) or 10 (c, f, i) min after treatments (×20). The experiments were repeated 2–3 times, with similar trends observed.