Synapse loss induced by interleukin-1β requires pre- and post-synaptic mechanisms

Abstract

Interleukin-1β (IL-1β) is an inflammatory cytokine that exerts marked effects on neuronal function and survival. Here we examined the effects of IL-1β on synapses between rat hippocampal neurons in culture using an imaging-based assay to quantify clusters of the scaffolding protein postsynaptic density 95 fused to green fluorescent protein. Treatment with IL-1β for 24 h induced a 23 ± 3% loss in the number of synaptic sites. Pharmacological studies indicated that synapse loss was mediated by the IL-1 receptor with subsequent activation of two pathways. COX2-mediated prostaglandin production and postsynaptic activation of a Src family tyrosine kinase were required. Presynaptic release of glutamate with subsequent activation of NMDA receptors was necessary for IL-1β-induced synapse loss. Neither Src activation nor prostaglandin E2 (PGE2) application alone was sufficient to reduce the number of synapses. However, in cells expressing constitutively active or pharmacologically activated Src, PGE2 induced synapse loss. Thus, IL-1β reduces the number of synaptic connections by simultaneously activating multiple pathways that require both pre- and post-synaptic activity. These results highlight targets that may prove important for pharmacotherapy of neuroinflammatory disease.

Fig. 1

Activation of IL-1β receptors induces synapse loss via NMDA receptors. a Confocal fluorescent images display maximum z-projections of neurons expressing PSD95-GFP and DsRed2. Processing of PSD95-GFP images identified PSDs as fluorescent puncta meeting intensity and size criteria and in contact with a mask derived from the DsRed2 image. The images in (a) were processed to produce the 0 h image in (b). b–d, labeled PSDs were dilated and overlaid on the DsRed2 image for visualization purposes. Processed images show cells before (0 h) and 24 h after no treatment (b control), treatment for 24 h with 3 ng/ml IL-1β c, and treatment with 10 μM MK801 15 min prior to and during 24 h treatment with IL-1β d. e Bar graph summarizes the effects of IL-1β on changes in PSD95-GFP puncta after 24 h treatment under control conditions (control, open bars) or following treatment with 3 ng/ml IL-1β (solid bars). Cells were untreated (n=45) or treated with 1 μg/ml IL-1ra (n=12), 10 μM MK801 (n=13) or 300 μM AP5 (n=6) as indicated. Data are mean ± SEM. *p<0.05 relative to control (Student’s t test); #, p<0.05 relative to IL-1β untreated (ANOVA with Bonferroni post-test).

Fig. 2

IL-1β requires glutamate release to evoke synapse loss. a–b Bar graphs summarize the effects of IL-1β on changes in PSD95-GFP (a) or Syn-GFP (b) puncta after 24 h treatment under control conditions (control, open bars) or following treatment with 3 ng/ml IL-1β (solid bars). a Cells expressing PSD95-GFP were untreated (n=45) or pretreated with 1 μM ω-conotoxin MVIIC (ω-CgTx; n=9) for 15 min or 0.3 μg/ml tetanus toxin (TetTx; n=5) for 24 h as indicated. b Cells expressing Syn-GFP were untreated (n=21) or treated with 1 μg/ml IL-1ra (n=10) or 10 μM MK801 (n=11) as indicated. Data are mean ± SEM. *p<0.05, **p<0.01 relative to control (Student’s t test); #, p<0.05, ### p<0.001 relative to IL-1β untreated (ANOVA with Bonferroni post-test).

Fig. 3

IL-1β evokes synapse loss via activation of a Src family kinase. a–b Images displaying PSDs were processed as described in Methods. Images show PSDs before (0 h) and 24 h after treatment with 3 ng/ml IL-1β in the presence of 10 μM PP2 (a) or PP3 (b). c Bar graph summarizes the effects of IL-1β on changes in PSD95-GFP puncta after 24 h treatment under control conditions (control, open bars) or following treatment with 3 ng/ml IL-1β (solid bars). Cells were untreated (n=45) or treated with 10 μM PP2 (n=14) or 10 μM PP3 (n=6) as indicated. Data are mean ± SEM. *p<0.05 relative to control (Student’s t test); #, p<0.05 relative to IL-1β untreated (ANOVA with Bonferroni post-test).

Fig. 4

IL-1β induced synapse loss requires prostaglandin production. Bar graph summarizes the effects of IL-1β on changes in PSD95-GFP puncta after 24 h treatment under control conditions (control, open bars) or following treatment with 3 ng/ml IL-1β (solid bars). Cells were untreated (n=45) or treated with 20 μM NS398 (n=9), 50 μM SB203580 (n=7), 10 μM AH6809 (n=10) or 10 μM AH23848 (n=11) as indicated. Data are mean ± SEM. *p<0.05 relative to control (Student’s t test); #, p<0.05 relative to IL-1β untreated (ANOVA with Bonferroni post-test).

Fig. 5

Synapse loss requires coincident activation of the EP2 receptor and Src kinase. a Bar graph summarizes the changes in synapse number in cells transfected with empty vector (open bars) or a CA-Src expression plasmid (solid bars) following 24 h exposure to no treatment (untreated, n=9) or treatment with 10 μM PGE2 (n=11). *p<0.05 relative to empty vector (Student’s t test); #, p<0.05 relative to untreated CA-Src (ANOVA with Bonferroni post-test). b Bar graph summarizes the changes in synapse number in cells under control conditions (control, open bars) or following treatment with 10 μM C2-ceramide (solid bars) following 24 h exposure to no additional treatment (untreated, n=4) or treatment with 10 μM PGE2 (n=13). *p<0.05 relative to control (Student’s t test); #, p<0.05 relative to C2-ceramide untreated (ANOVA with Bonferroni post-test).

Fig. 6

Schematic displays hypothesized pre- and post-synaptic signaling pathways activated by IL-1β to induce synapse loss. Blunt lines and arrows, respectively denote conclusions based on the inhibitory and activating treatments tested in this study. Aspects of the model based on published reports are labeled numerically. 1) IL-1R activation results in the activation of Src family kinases (SFK) (Viviani et al., 2006;(Weber et al., 2010). 2) SFKs phosphorylate and increase the sensitivity of NMDA receptors to glutamate (Salter et al., 2009). 3) Increased Ca²⁺ influx via NMDA receptors activates the ubiquitin-proteasome pathway to produce synapse loss (Kim et al., 2008; Waataja et al., 2008). IL1-R activation in neurons (4) (Samad et al., 2001) and astrocytes (5) (Hartung et al., 1989) leads to p38 MAPK mediated transcription of the COX2 gene. 6) PGE2 activation of presynaptic EP2 receptors increases the release of glutamate (Sang et al., 2005).