GPR17 agonist specificity in [³⁵S]GTPγS binding. (A, B) Agonist-response curves to CysLTs and nucleotides in 1321N1 cells expressing hGPR17 (A), as shown by the presence of a specific 1087 bp amplification product (B). No products were detected in the absence of retrotranscription (RT) (indicated as –RT), nor in cells transfected with the empty vector. (C, D), same as (A) and (B) for rGPR17 (RT–PCR product of 1079 bp). (E) Responses to CysLTs in COS-7 cells expressing hGPR17. (F) Responses to nucleotides in HEK-293 cells expressing hGPR17. Each point is the mean±s.d. of 4–9 independent experiments run in triplicate. For each agonist, EC₅₀ values are also reported.

A key role for Gi proteins and adenylyl cyclase in GPR17 function. (A) Pre-incubation of 1321N1 cells expressing the human receptor with the Gi-protein inhibitor PTX (black columns) significantly reduced stimulation of [³⁵S]GTPγS binding by the indicated agonists with respect to untreated cells (white columns) (P<0.001). Data are the mean of three experiments run in triplicate. (B, C) Effect of uracil nucleotides and LTD₄ on FK-stimulated cAMP accumulation. 1321N1 cells transfected with either hGPR17 (B) or with corresponding empty plasmid (C) were treated with the indicated agonists in the presence of 10 μM FK. cAMP levels (means±s.e.m.; n=6) were quantified and expressed as percentage of cAMP in the presence of FK alone set to 100% (white bar in graphs). A concentration-dependent inhibition of FK-stimulated cAMP production was detected only in hGPR17-expressing cells. (D) Preincubation of 1321N1 cells expressing hGPR17 with PTX under the same conditions reported in (A) before agonist treatment fully obliterated the effects of uracil nucleotides and LTD₄ on FK-stimulated cAMP formation. Basal and FK-stimulated cAMP levels were 11.5±1.2 and 89.5±6.5 pmol/ml, respectively (values refer to 10 μg of protein/sample and a 10 min incubation of intact cells with either medium alone or medium+FK).

GPR17 immunoreactivity in transfected cells and in rat brain. (A–C) 1321N1 cells transfected with empty vector (A, control cells) or rGPR17-pcDNA3.1 (B, transfected cells) were incubated with anti-GPR17 antisera in the absence or presence (C) of the peptide used for rabbit immunization (neutralizing peptide). (D–F) Same as in (A–C) on COS-7 cells. (G–I) Double-labeling of rat cortical slices with anti-GPR17 antibody (green fluorescence) and astroglial (G, GFAP) or neuronal markers (H, SMI-311 or I, β-III tubulin) (red fluorescence). (J–N) At 48 h after MCAo, rat brain slices were incubated with anti-GPR17 antisera followed by either horseradish peroxidase staining (J–M) or double-staining with anti-SMI-311 (N). Micrographs were taken inside the ischemic area (J, gray area in left lesioned hemisphere) or at its borders (K), or in the corresponding healthy brain areas of the contralateral unlesioned right hemisphere (L, M), as indicated by squares in the drawing of coronal brain section. Inside the ischemic infarct, GPR17 immunoreactivity was found on pyramidal-like cells (higher magnification inset in J), which were also positive for SMI-311 (N). Similar data were obtained in five experiments (24 brain sections in total). Scale bars 30 μm.

Sequence analysis of GPR17 and expression of the human and rat receptors. (A) Phylogenetic tree showing the relationships between GPR17 and P2Y and CysLT receptors. (B) Alignment of human, mouse and rat amino-acid GPR17 sequences highlighting the seven TM domains and the conserved H-X-X-R motif in TM6. (C) RT–PCR amplification of the human (1087 bp) or rat (1079 bp) cDNA sequences in brain, kidney, heart and in human umbilical vein endothelial cells (HUVEC) and breast adenocarcinoma MCF7 cells. No signal was found in human peripheral blood mononuclear cells (PBMC), bronchial smooth muscle cells (bSMC), myeloid U937, hepatocellular carcinoma Hep-G2, cervix carcinoma (HeLa) and neuroblastoma SK-N-BE cells. Parallel expression of the housekeeping gene beta-actin is shown.

Effect of P2Y and CysLT₁ receptor antagonists on the activation of recombinant GPR17 in [³⁵S]GTPγS binding. (A) Antagonism of UDP-glucose stimulation of [³⁵S]GTPγS binding by the indicated P2Y antagonists in 1321N1 cells expressing hGPR17. (B) Same as in (A), in 1321N1 cells expressing rGPR17. (C) Antagonism of LTD₄ stimulation of [³⁵S]GTPγS binding by indicated CysLT₁ antagonists in 1321N1 cells expressing hGPR17. (D) Same as in (C), in 1321N1 cells expressing rGPR17. Flat concentration–response curves in (A–D) refer to cells transfected with the pcDNA3.1 empty vector. Each point is expressed as percentage of basal [³⁵S]GTPγS-specific binding set to 100% and is the mean±s.d. of 4–7 independent experiments run in triplicate. For each antagonist, IC₅₀ values are reported.

Single-cell calcium imaging in cells expressing hGPR17. Each trace shows response recorded from one single cell. (A–C) Approximately 30% of 1321N1 cells expressing hGPR17 showed responses to uracil nucleotides (mean calcium response to UDP-glucose: ΔF_340/380=0.23±0.06, mean±s.e.m., n=15; mean calcium response to UDP: ΔF_340/380=0.12±0.01, n=11; mean calcium response to UDP-galactose: ΔF_340/380=0.05±0.003, n=3) or (D) to LTD₄ (ΔF_340/380=0.26±0.05, mean±s.e.m., n=18). (E–H) The same agonists induced no responses in cells transfected with the empty plasmid. (I) Approximately 45% of COS-7 cells transfected with hCysLT₁ receptor showed calcium transients to LTD₄ (mean calcium response: ΔF_340/380=0.3±0.08, mean±s.e.m., n=17). (J) Similar responses were recorded from approximately 50% of COS-7 cells expressing hGPR17 (mean calcium response: ΔF_340/380=0.18±0.03, mean±s.e.m., n=22). (K) No responses to LTD₄ were recorded in cells transfected with the empty plasmid.

Effect of montelukast (Mtk), cangrelor (Cang) and oligo616 on evolution of brain infarct size as determined by MRI at 2, 24 and 48 h after MCAo. (A) Representative Tr(D) images of coronal brain sections from ischemic rats treated with either Mtk, Cang, oligo616 (616) or scrambled oligonucleotide (Scr) in comparison with corresponding control animals treated with vehicle only (C1, C2, C3). Antisense oligonucleotides were administered 48 and 24 h before and 10 min after MCAo (indicated as ‘pre'); in some experiments, animals only received a single dose of antisense oligonucleotide (indicated as ‘post', also see below). In left lesioned hemispheres, ischemia-associated brain damage is shown by black areas. (B) Quantitative analysis of infarct size volume at 24 and 48 h after MCAo from rats receiving vehicle (C1, n=5) or Mtk (2 mg/kg intravenously (i.v.), 10 min after MCAo; n=6); vehicle (C2, n=5) or Cang (4.5 μg/animal, i.c.v., 10 min after MCAo, n=5), vehicle (C3, n=5) or either oligo616 or scrambled-oligo (400 ng/animal, i.c.v., 48 and 24 h before and 10 min after MCAo, n=5, indicated as ‘pre'). Some animals received a single i.c.v. dose of 4.5 μg of either oligo616 (n=6) or scrambled-oligo (n=6) 10 min after ischemia (indicated as ‘post'). Data are expressed as percentage variation of infarct volume at 24 and 48 h after MCAo compared to 2 h considered as 100%. ^§P<0.05, ^§§P<0.01 versus 2 h; ^*P<0.05, ^**P<0.01 versus corresponding control animals; ^#P<0.05, ^##P<0.01 versus corresponding Scr animals.