G-protein coupled receptor-evoked glutamate exocytosis from astrocytes: role of prostaglandins

Abstract

Astrocytes are highly secretory cells, participating in rapid brain communication by releasing glutamate. Recent evidences have suggested that this process is largely mediated by Ca(2+)-dependent regulated exocytosis of VGLUT-positive vesicles. Here by taking advantage of VGLUT1-pHluorin and TIRF illumination, we characterized mechanisms of glutamate exocytosis evoked by endogenous transmitters (glutamate and ATP), which are known to stimulate Ca(2+) elevations in astrocytes. At first we characterized the VGLUT1-pHluorin expressing vesicles and found that VGLUT1-positive vesicles were a specific population of small synaptic-like microvesicles containing glutamate but which do not express VGLUT2. Endogenous mediators evoked a burst of exocytosis through activation of G-protein coupled receptors. Subsequent glutamate exocytosis was reduced by about 80% upon pharmacological blockade of the prostaglandin-forming enzyme, cyclooxygenase. On the other hand, receptor stimulation was accompanied by extracellular release of prostaglandin E2 (PGE2). Interestingly, administration of exogenous PGE2 produced per se rapid, store-dependent burst exocytosis of glutamatergic vesicles in astrocytes. Finally, when PGE2-neutralizing antibody was added to cell medium, transmitter-evoked exocytosis was again significantly reduced (by about 50%). Overall these data indicate that cyclooxygenase products are responsible for a major component of glutamate exocytosis in astrocytes and that large part of such component is sustained by autocrine/paracrine action of PGE2.

Figure 1

VGLUT1-pHluorin is mainly expressed on a specific population of glutamatergic synaptic like microvesicles. In the figure the left panels (in green) show astrocytes transfected with VGLUT1-pHluorin construct revealed by rabbit antibody against GFP. The middle panels (in red) show the markers of the intracellular secretory organelles, revealed by mouse antibodies against specific markers of ((a)–(e)) synaptic like microvesicles ((a) VGLUT1, (b) VGLUT2, (c) VAMP3, (d) glutamate, (e) VAMP2), of (f) dense core granules (phogrin), of (g) late endosomes, multivesicular bodies and lysosomes (LAMP1), of (h) early endosomes (EAA1), and of (i) recycling endosomes (transferrin receptor, Tfr). The right panels show the merged images. Bars: 20 μm.

Figure 2

Analysis of VGLUT1-pHluorin vesicles that colocalize with markers of early or recycling endosomes. (a)–(d) Estimation of the size of vesicles expressing VGLUT1-pHluorin. Analysis of individual vesicle was performed in confocal images of VGLUT1-pHluorin-expressing astrocytes by plotting fluorescence intensity of pHluorin spots against distance from the centre of the spot (black curve ± SD). Such an analysis provided an estimation of the average fluorescence profile otherwise called “radial sweep” [23]. The fluorescence intensity values obtained from the radial sweep were well fitted by a one-dimensional Gaussian function (red curve). Such a curve represents the average radial sweep value obtained from 20 vesicles. Note that the half maximum value of pHluorin-expressing vesicle positive for EAA1 ((b), marker of early endosomes, 490 ± 5 nm) is similar to that of 200 nm fluorescent beads ((a), 506 ± 6 nm) and the half maximum value of pHluorin-expressing vesicle that do not express EAA1 ((d), 349 ± 7 nm) is similar to that of 40 nm fluorescent beads ((c), 361 ± 6 nm). (e), (f) Temporal distribution of VGLUT1-pHluorin and Alexa-Tf 568 fusion events evoked by DHPG application. (e) Each individual histogram represents the number (mean ± SD) of fusion events detected from VGLUT1-pHluorin vesicles in a 50 ms-long frame (n = 5 cells). (f) Fusion events (mean ± SD) detected from VGLUT1-pHluorin and Alexa-Tf568 double positive vesicles in the same cells as in (e). Each histogram represents the number of fusion events detected in a 50 ms-long frame (n = 5 cells).

Figure 3

Pharmacological characterization of the receptor subtypes mediating exocytosis of VGLUT1-pHluorin positive vesicles in response to ATP and glutamate agonists. (a) TIRF image showing an astrocyte transfected with VGLUT1-pHluorin. Bar 20 nm. (b) Stereotyped sequence of pHluorin destaining reveals exocytosis of a VGLUT1-pHluorin positive vesicle. The sequential gray scale micrographs represent the fate of pHluorin before (−100 ms) and during (100, 200, 400 ms) the fusion event. Bars: 380 nm. The scheme shows the behaviour of pHluorin before and after fusion event. Note that the color code for the pHluorin fluorescence signal is gray when the signal is off and green when it is on. (c), (d) P2Y₁ receptors mediate the ATP-evoked exocytosis. (c) Temporal distribution of fusion events evoked by ATP (100 μM). (d) Histograms represent the total number of fusion events evoked by ATP (417.14 ± 32.4) that is strongly inhibited in the presence of the P2 purine antagonists PPADS (100 μM, 58.6 ± 7) as well as of the P2Y₁-selective compound, A3P5PS (100 μM, 70.2 ± 5.8). Data are ± SEM of 4 cells. (e), (f) mGluR5 mediates the response to t-ACPD, in the presence of AMPA. (e) Temporal distribution of fusion events evoked by 50 μM t-ACPD+50 μM AMPA. (f) Histograms represent the total number of fusion events evoked by t-ACPD+AMPA (447.1 ± 28.7) that is strongly inhibited in the presence of the mGluR antagonists, including the subtype-nonselective MCPG (500 μM, 98.3 ± 7.4) and the mGluR5-selective MPEP (200 nM, 80.1 ± 7). Data are ± SEM of 4 cells. Statistical significance of inhibition with receptor antagonists was calculated using t-test (**P < 0.01).

Figure 4

COX blockers strongly inhibit the exocytosis of glutamate evoked by activation of group I mGluR and of purinergic P2Y₁ receptor. (a), (b) Quantitative histograms represent the total number of fusion events evoked by either DHPG (100 μM, 456.7 ± 54.8) or 2MeSADP (20 μM, 467.6 ± 46.8) in the presence of COX blockers, INDO (1 μM; 94.9 ± 9.6, 88.8 ± 5.4, resp.) or ASA (10 μM, 86.4 ± 11.7, 79.5 ± 14.4, resp.). Data are ± SEM of 4 cells. Statistical significance of inhibition with receptor antagonists was calculated using t-test (**P < 0.01).

Figure 5

Extracellular PGE₂: accumulation in response to various stimuli and effects on exocytosis of glutamatergic vesicles. (a) Extracellular accumulation of PGE₂ (expressed as pg/mL) in response to 3 min stimulation with either t-ACPD+ AMPA (each at 50 μM) or ATP (100 μM). Each point represents the average ± SEM of two experiments in triplicate with each stimulus. (b) Temporal distribution of fusion events evoked by PGE₂ (50 μM). Inset histograms represent the total number of fusion events evoked by PGE₂ (349 ± 26) in the presence of 0 mM Ca²⁺ and 5 mM of EGTA (345 ± 32) or cyclopiazonic acid (CPA, 10 μM, 25 ± 12). (c) Temporal distribution of fusion events evoked by DHPG (100 μM). (d) Inhibitory effect of AbPGE₂ (buffering capacity >1000 pg/mL PGE₂) on exocytosis of glutamatergic vesicles evoked by DHPG (100 μM). Histograms represent temporal distribution of fusion events evoked by DHPG in the presence of AbPGE₂. Statistical significance was calculated using t-test (*P < 0.05).