Pathway specific modulation of S1P1 receptor signalling in rat and human astrocytes

Abstract

Background and purpose: The sphingosine 1-phosphate receptor subtype 1 (S1P1R) is modulated by phosphorylated FTY720 (pFTY720), which causes S1P1R internalization preventing lymphocyte migration thus limiting autoimmune response. Studies indicate that internalized S1P1Rs continue to signal, maintaining an inhibition of cAMP, thus raising question whether the effects of pFTY720 are due to transient initial agonism, functional antagonism and/or continued signalling. To further investigate this, the current study first determined if continued S1P1R activation is pathway specific.

Experimental approach: Using human and rat astrocyte cultures, the effects of S1P1R activation on cAMP, pERK and Ca(2+) signalling was investigated. In addition, to examine the role of S1P1R redistribution on these events, a novel biologic (MNP301) that prevented pFTY720-mediated S1P1R redistribution was engineered.

Key results: The data showed that pFTY720 induced long-lasting S1P1R redistribution and continued cAMP signalling in rat astrocytes. In contrast, pFTY720 induced a transient increase of Ca(2+) in astrocytes and subsequent antagonism of Ca(2+) signalling. Notably, while leaving pFTY720-induced cAMP signalling intact, the novel MNP301 peptide attenuated S1P1R-mediated Ca(2+) and pERK signalling in cultured rat astrocytes.

Conclusions and implications: These findings suggested that pFTY720 causes continued cAMP signalling that is not dependent on S1P1R redistribution and induces functional antagonism of Ca(2+) signalling after transient stimulation. To our knowledge, this is the first report demonstrating that pFTY720 causes continued signalling in one pathway (cAMP) versus functional antagonism of another pathway (Ca(2+)) and which also suggests that redistributed S1P1Rs may have differing signalling properties from those expressed at the surface.

Figure 1

pFTY720 treatment causes redistribution of S1P1R in a time and concentration dependent manner in astrocytes. (A) Pure rat astrocytes were stained for GFAP, neurofilament H, CD11b and CNPase. A total of 24 images were analysed (six images per group). Average percentage of positively stained cells for each group was as follows: GFAP 98.58% ± 0.57, CD11b 1.35% ± 0.77 and CNPase 1.83% ± 0.48. No neurofilament H positive cells were observed. Scale bars 50 μm. Pure astrocyte cultures were treated with (B) increasing concentrations of pFTY720 for 1 h or (C) with 1 μM pFTY720 for time periods indicated. Unless indicated, cells were immunostained with GFAP Ab (red) and S1P1R Ab (green), cell nuclei appear as blue (Hoescht) and arrows indicate areas of S1P1R redistribution as determined by perinuclear staining. Scale bars 20 μm.

Figure 2

pFTY720 induced long lasting redistribution of S1P1R to the trans-Golgi-network in astrocytes. (A) DMSO (vehicle control), 1 μM pFTY720 treatment for 1 h (positive control) and 1 μM pFTY720 treatment for 1 h followed by a 5-h washout period. Scale bars 20 μm. (B) Calculation of S1P1R redistribution was performed by taking mean fluorescent measurements from the ‘perinuclear zone’ defined as S1P1R staining <10 μm from the nucleus (inner circle) and the ‘process zone’ defined as S1P1R staining >35 μm from the nucleus (outer circle). (C) Bar graphs show changes in the levels of perinuclear and cytoplasmic S1P1R staining in astrocytes treated with pFTY720. Statistical analysis shows the level of redistribution and also the ratio of nuclear : cytoplasmic S1P1R expression (***P < 0.001 vs. control, two-way anova and Bonferroni post hoc test). (D) Vehicle control and pFTY720 treated astrocytes were stained with organelle marker antibodies (red) including early endosome (EEA-1), trans-Golgi-network (p230), lysosome (LAMP1) and Golgi matrix (GM130). S1P1R staining overlapped with p230 and GM130 markers indicative of localization of the redistributed S1P1R to the TGN and Golgi matrix. (E) S1P1R redistribution was investigated following treatment with DMSO (Ctrl), pFTY720 (1 μM, 1 h), the selective S1P1 agonist SEW2871 (1 μM, 1 h) and the endogenous ligand S1P (1 μM, 1 h). Scale bars 50 μm. Bar graph shows statistical analysis of redistribution (***P < 0.001 vs. control, one-way anova and Bonferroni post hoc test). Unless indicated, cells were immunostained with GFAP Ab (red) and S1P1R Ab (green), cell nuclei appear as blue (Hoescht) and arrows indicate areas of S1P1R redistribution as determined by perinuclear staining.

Figure 3

MNP301 prevents pFTY720-induced redistribution of S1P1R. (A) The structure of MNP301 is shown and is composed of a FITC tag, a cell transduction Tat sequence and the last 10 residues of ct-S1PR (FITC-Ahx-YGRKKRRQRRR-MSSGNVNSSS). (B) Pure astrocytes cultures were treated with increasing concentrations of FITC-Tat-MNP301 for 1 h at 37°C and direct FITC fluorescence (green) was observed at a wavelength of 488 nm by confocal microscopy. Graph shows mean fluorescence calculated from five images per condition. Significant cellular transduction was observed following incubation of astrocytes with 100 μg·mL⁻¹ MNP301 (***P < 0.001 vs. control, one-way anova and Bonferroni post hoc test). (C) Organotypic cerebellar culture were treated for 2 h with FITC-Tat-MNP301 (250 μg·mL⁻¹) and placed in fresh medium for 2 days. Slices were stained for S1P1R Ab (purple) and GFAP Ab (grey) 2 days post-tansduction with MNP301 (green). Boxed inset shows magnification of white box of FITC-Tat-MNP301 fluorescence. (D) Astrocytes were treated with pFTY720 (1 μM) for 1 h at 37°C in the presence and absence of a non-FITC labelled, Tat-fused MNP301 (100 μg·mL⁻¹). Scale bars 50 μm. (E) Relative cumulative frequency distribution curves, as determined by automated image analysis, display a marked increase (>20% of total cells analysed) in the number of astrocytes exhibiting S1P1R redistribution in the pFTY720-treated group; as measured by the ratio of perinuclear : cytoplasmic fluorescence. Co-treatment with MNP301 attenuated this rightward distribution shift. The figures show no difference in S1P1R redistribution between (i) control vs. MNP301 or (ii) MNP301 vs. MNP301 + pFTY720, and a difference of S1P1R redistribution between (iii) pFTY720 vs. MNP301 and (iv) pFTY720 vs. MNP301 + pFTY720. The number of cells analysed per treatment group were vehicle control = 154; MNP301 = 144; pFTY720 = 225; MNP301 + pFTY720 = 264. (F) Statistical analysis of S1P1R redistribution, as derived from manual image analysis, indicates that MNP301 significantly inhibited pFTY720 induced redistribution to the TGN. Data expressed as mean ± SEM of three separate experiments (***P < 0.001 vs. pFTY720 alone, one-way anova, Bonferroni post hoc test). (G) Treatment of rat astrocytes for 1 h at 37°C with 100 μg·mL⁻¹ control peptide (non-FITC labelled, Tat-fused Ctrl-pep; YGRKKRRQRRR-VCMGDHWFDV) did not alter S1P1R localization in the presence or absence of pFTY720 (1 μM for 1 h at 37°C). Unless indicated, cells were immunostained with GFAP Ab (red) and S1P1R Ab (green), cell nuclei appear as blue (Hoescht) and arrows indicate areas of redistribution as determined by perinuclear staining. (H) Human astrocytes were pretreated with or without MNP301 (100 μg·mL⁻¹ for 1 h) followed by treatment with or without pFTY720 (1 μM for 1 h). Astrocytes were then incubated with non-permeable biotin (0.5 mg·mL⁻¹ for 30 min). The cell surface biotin proteins were immunoprecipitated using streptavidin-coated agarose beads and levels of cell surface biotinylated S1P1R were measured by Western blotting. Data shown is a representative of three independent experiments.

Figure 4

Continued pFTY720-induced cAMP signalling of S1P1R in astrocytes. (A) Shown are the concentration-response curves of acute pFTY720 (1 μM for 20 min) treatment inhibiting forskolin-induced cAMP levels in astrocytes pretreated with or without MNP301 (±100 μg·mL⁻¹ for 1 h). Data presented as percentage cAMP inhibition and is representative of three separate experiments. (B) The acute effect of pFTY720 (1 μM for 20 min) on forskolin-induced cAMP formation in astrocytes is shown (lane 2). The percentage of cAMP inhibition in astrocytes pre-incubated without (lane 3) or with (lane 4) MNP301 (100 μg·mL⁻¹ for 1 h) followed by addition of pFTY720 (1 μM for 1 h) then washed and cAMP levels measured 5 h later is shown. Data are representative of five independent experiments. (***P < 0.001 vs. control, one-way anova, Bonferroni post hoc test). (C) Treatment of purified rat astrocyte cultures with pFTY720 (10 nM for 10 min) and S1P (100 nM for 10 min) induced a significant increase in the phosphorylation of ERK (pERK) as determined by Western blotting. Pretreatment with MNP301 (100 μg·mL⁻¹ for 1 h) did not significantly reduce the levels of pERK. (D) AUY954 (10 nM for 10 min) induced a significant increase in pERK in astrocytes that was significantly inhibited by pretreatment MNP301 (100 μg·mL⁻¹ for 1 h) (*P < 0.05, one-way anova with Neumann–Keuls post hoc test). (E) The significant increase in levels of pERK induced by glutamate (600 μM for 10 min) (*P < 0.05 based on one-way anova and Bonferroni post hoc analysis) were not significantly altered by MNP301 pretreatment (100 μg·mL⁻¹ for 1 h). (D-E) Data expressed as mean ± SEM of two–four separate experiments. The levels of phosphorylated ERK (P-ERK) were expressed as arbitrary units of optical density, following the correction for content of total ERK (T-ERK).

Figure 5

S1PR agonists induce a concentration dependent increase in Ca²⁺ levels in primary rat astrocytes. Increasing concentrations of (A) the natural ligand S1P and (B) the S1P1 specific agonist AUY954 induce a concentration dependent increase in Ca²⁺ levels. Representative images show time-lapse series, at basal levels time point 0 s (basal), after addition of 1 μM S1P or AUY954 at 30 s (S1P or AUY), and after stimulation with 30 μM glutamate at 180 s (Glu). Traces depict changes in Ca²⁺ levels over time. Bar graphs (30–90 s) show that S1P and AUY954 cause a concentration dependent increase in Ca²⁺ levels. Data represented as mean ± SEM. Cells counted, S1P 1 μM = 200, S1P 100 nM = 185, S1P 10 nM = 132; AUY 1 μM = 235, AUY 100 nM = 171, AUY 10 nM = 85.

Figure 6

S1PR pre-activation inhibits further S1PR mediated increases in Ca²⁺ levels in astrocytes. Pre-treatment with S1P, pFTY720 and AUY954 significantly inhibited changes in Ca²⁺ levels in response to a secondary treatment with (A) S1P or (B) AUY954. Representative images show time-lapse series, before addition of S1P or AUY954 (0 s, basal), after addition of 1 μM S1P or AUY954 (30 s) and after stimulation with 30 μM glutamate at 180 s (Glu). Traces depict changes in Ca²⁺ levels over time. Bar graphs (30–90 s) show data represented as mean ± SEM, with a total of eight observations for all conditions (***P < 0.001 pretreatment vs. control, as determined by unpaired t-test).

Figure 7

MNP301 antagonizes S1P1 receptor mediated increases in Ca²⁺ levels in astrocytes. Pretreatment with 100 μg·mL⁻¹ MNP301, but not the unrelated control peptide (Ctrl-pep), significantly inhibited changes in Ca²⁺ levels in response to a secondary treatment with (A) S1P or (B) AUY954. Representative images show time-lapse series, before addition of S1P or AUY954 (0 s, basal), after addition of 1 μM S1P or AUY954 (30 s) and after stimulation with 30 μM glutamate at 180 s (Glu). Traces depict changes in Ca²⁺ levels over time induced by (C) S1P and (D) AUY954. (E) Bar graph show data (30–90 s) represented as mean ± SEM (**P < 0.01 pretreatment vs. control, as determined one-way anova). Cells counted, Ctrl-S1P = 648, MNP301-S1P = 612; Ctrl-AUY = 479, MNP301-AUY = 399.

Figure 8

Agonism of S1PRs evokes increases in Ca²⁺ levels in human astrocytes from a predominantly extracellular source. (A) Human astrocyte cultures stained for GFAP (red) and S1P1R (green), show expression of S1P1 receptors in astrocytes. (B) Both S1P and the S1P1 specific compound SEW2871 induced a concentration dependent increase in Ca²⁺ levels in human astrocytes. Representative images show time-lapse series, after addition of increasing concentrations of S1P and SEW2871 (30 s). (C) Pretreatment of human astrocytes with 30 μM dantrolene (Dan), in the presence and absence of Ca²⁺ significantly reduced cells response to 1 μM SEW2871 (***P < 0.001 dantrolene vs. SEW2871 one-way anova, Dunnetts post hoc test). Pretreatment with 1 mM EGTA significantly inhibited 1 μM SEW2871 induced increases in Ca²⁺ levels (***P < 0.001 EGTA vs. SEW2871, unpaired t-test). Bar graphs (30–90 s) show averaged data ± SEM, n = 3 (***P < 0.001 vs. control, one-way anova and Bonferroni post hoc test). Scale bars 50 μm.