Microglial P2Y12 deficiency/inhibition protects against brain ischemia

Abstract

Objective: Microglia are among the first immune cells to respond to ischemic insults. Triggering of this inflammatory response may involve the microglial purinergic GPCR, P2Y12, activation via extracellular release of nucleotides from injured cells. It is also the inhibitory target of the widely used antiplatelet drug, clopidogrel. Thus, inhibiting this GPCR in microglia should inhibit microglial mediated neurotoxicity following ischemic brain injury.

Methods: Experimental cerebral ischemia was induced, in vitro with oxygen-glucose deprivation (OGD), or in vivo via bilateral common carotid artery occlusion (BCCAO). Genetic knock-down in vitro via siRNA, or in vivo P2Y12 transgenic mice (P2Y12-/- or P2Y12+/-), or in vivo treatment with clopidogrel, were used to manipulate the receptor. Neuron death, microglial activation, and microglial migration were assessed.

Results: The addition of microglia to neuron-astrocyte cultures increases neurotoxicity following OGD, which is mitigated by microglial P2Y12 deficiency (P<0.05). Wildtype microglia form clusters around these neurons following injury, which is also prevented in P2Y12 deficient microglia (P<0.01). P2Y12 knock-out microglia migrated less than WT controls in response to OGD-conditioned neuronal supernatant. P2Y12 (+/-) or clopidogrel treated mice subjected to global cerebral ischemia suffered less neuronal injury (P<0.01, P<0.001) compared to wild-type littermates or placebo treated controls. There were also fewer microglia surrounding areas of injury, and less activation of the pro-inflammatory transcription factor, nuclear factor Kappa B (NFkB).

Interpretation: P2Y12 participates in ischemia related inflammation by mediating microglial migration and potentiation of neurotoxicity. These data also suggest an additional anti-inflammatory, neuroprotective benefit of clopidogrel.

Figure 1. Microglia potentiate neuron death following in vitro ischemia.

Representative images of mixed neuron astrocyte (NA) cultures prior to oxygen glucose deprivation (OGD) (A, NA con) and after (B, NA OGD). Primary microglial cells added to NA cultures are visible as small phase bright cells (C. NAM con). Following OGD, the previously confluent NA cell layer is disrupted leaving mostly microglial cells which tend to form clusters (D, NAM OGD). E: Mixed cultures of primary neurons and astrocytes (NA) were prepared and either primary microglia (NAM) or BV2 cells were added (NAB). Neuron death due to OGD was increased in mixed cultures containing either primary microglia (NAM) or BV2 cells (NAB), and the extent of death was similar. Microglia appear to require cell-cell contact, since BV2 cells did not potentiate neurotoxicity when cultured in inserts (NABi). (*P<0.001 vs. NA, NABi).

Figure 2. P2Y₁₂ deficiency in microglia improved neuron survival following OGD and prevented microglial clustering.

P2Y₁₂ was knocked down in BV2 cells using siRNA, and were cocultured with mixed neuron-astrocyte cultures. Scrambled siRNA was used as a control. Following OGD, neuron viability was decreased in BV2 cells receiving control siRNA, but was increased where P2Y₁₂ was deficient (A). The tendency towards microglial clustering was increased by OGD, but prevented by P2Y₁₂ deficiency (B, C). Immunostains of the microglial marker isolectin B4 (IB4) indicate that the majority of the cells forming the clusters are indeed microglia (C). **P<0.01.

Figure 3. Microglia migrate in response to OGD injured neurons, but not astrocytes. Migration is inhibited under conditions of P2Y₁₂ deficiency.

A: A Boyden-like assay was used to estimate the ability of BV2 cells to travel through a cell culture insert. BV2 cells were placed in inserts, which were then placed above wells containing various potential chemotactic stimulants. Cells migrating into the lower chamber were then quantified using a fluorescent detection dye. Test media are indicated in the upper row of the x-axis label (media = BSS 5.5; AST = astrocyte-conditioned media; Neu = neuron-conditioned media). Cultures previously exposed to OGD are indicated by ‘+’, and those not exposed by ‘−’. OGD increased BV2 cell migration neuron conditioned media, but not astrocyte conditioned media. B: Using a Dunn chamber assay where distance traveled can be estimated, primary microglia harvested from wildtype (WT) or P2Y₁₂ deficient (P2Y₁₂ ^−/−, KO) mice, and were exposed to conditioned media from OGD-exposed wildtype neurons. The distance traveled over a 30 minute observation period is shown. Primary microglia from KO mice traveled shorter distances than WT microglia. (*P<0.05, **P<0.01).

Figure 4. Mice deficient in P2Y₁₂ are protected from global cerebral ischemia and also have decreased microglia surrounding areas of injury.

Following bilateral common carotid artery occlusion for 12 min followed by 3 d reperfusion, P2Y₁₂ deficient mice (P2Y₁₂ ^+/−) suffered less injury to hippocampal CA1 compared to wildtype (WT) mice. WT mice treated with clopidogrel (WT+CL, Clopidogrel) were also protected (A, B). Following BCCAO, microglial densities as estimated from counts of isolectin B4 (IB4) positive cells were also decreased within CA1 among P2Y₁₂ ^+/− and Clopidogrel treated mice (A, C). *P<0.05, **P<0.01.

Figure 5. P2Y₁₂ deficiency or inhibition decreases NFκB expression.

Numbers of cells positive for NFκB’s p65 subunit were decreased among P2Y₁₂ ^+/− and treated mice (A, C). Furthermore, cells of P2Y₁₂ ^+/− and treated mice had less nuclear NFκB staining (A, green arrows) compared to untreated WT following BCCAO (A, D). B: Double labeling for NFκB’s p65 subunit (red) and the microglial marker CD11b (green) show that most CD11b cells are also positive for NFκB after BCCAO in WT brains. Several microglia in brains of P2Y₁₂ ^+/− mice or WT mice treated with clopidogrel (WT+CL) are not NFκB positive (arrows). *P<0.05, **P<0.01, scale bar = 25 µm.