Interferon-gamma increases neuronal death in response to amyloid-beta1-42

Abstract

Background: Alzheimer's disease is a neurodegenerative disorder characterized by a progressive cognitive impairment, the consequence of neuronal dysfunction and ultimately the death of neurons. The amyloid hypothesis proposes that neuronal damage results from the accumulation of insoluble, hydrophobic, fibrillar peptides such as amyloid-beta1-42. These peptides activate enzymes resulting in a cascade of second messengers including prostaglandins and platelet-activating factor. Apoptosis of neurons is thought to follow as a consequence of the uncontrolled release of second messengers. Biochemical, histopathological and genetic studies suggest that pro-inflammatory cytokines play a role in neurodegeneration during Alzheimer's disease. In the current study we examined the effects of interferon (IFN)-gamma, tumour necrosis factor (TNF)alpha, interleukin (IL)-1beta and IL-6 on neurons.

Methods: Primary murine cortical or cerebellar neurons, or human SH-SY5Y neuroblastoma cells, were grown in vitro. Neurons were treated with cytokines prior to incubation with different neuronal insults. Cell survival, caspase-3 activity (a measure of apoptosis) and prostaglandin production were measured. Immunoblots were used to determine the effects of cytokines on the levels of cytoplasmic phospholipase A2 or phospholipase C gamma-1.

Results: While none of the cytokines tested were directly neurotoxic, pre-treatment with IFN-gamma sensitised neurons to the toxic effects of amyloid-beta1-42 or HuPrP82-146 (a neurotoxic peptide found in prion diseases). The effects of IFN-gamma were seen on cortical and cerebellar neurons, and on SH-SY5Y neuroblastoma cells. However, pre-treatment with IFN-gamma did not affect the sensitivity to neurons treated with staurosporine or hydrogen peroxide. Pre-treatment with IFN-gamma increased the levels of cytoplasmic phospholipase A2 in SH-SY5Y cells and increased prostaglandin E2 production in response to amyloid-beta1-42.

Conclusion: Treatment of neuronal cells with IFN-gamma increased neuronal death in response to amyloid-beta1-42 or HuPrP82-146. IFN-gamma increased the levels of cytoplasmic phospholipase A2 in cultured neuronal cells and increased expression of cytoplasmic phospholipase A2 was associated with increased production of prostaglandin E2 in response to amyloid-beta1-42 or HuPrP82-146. Such observations suggest that IFN-gamma produced within the brain may increase neuronal loss in Alzheimer's disease.

Figure 1

Pre-treatment with IFN-γ reduces the survival of neurons incubated with amyloid-β_1-42. The survival of primary cortical neurons pre-treated with 1 ng/ml TNFα, IL-1β, IL-6 or IFN-γ prior to the addition of 10 μm amyloid-β_1-42 (shaded bars) or 10 μm amyloid-β_42-1 (open bars). Values shown are the mean percentage cell survival from triplicate experiments repeated 3 times (n = 9), ± standard deviation (SD). ** = Neuronal survival significantly less than untreated neurons incubated with amyloid-β_1-42 (p < 0.05).

Figure 2

The IFN-γ-induced sensitization of neurons to amyloid-β_1-42 is dose-dependent. The survival of primary cortical neurons pre-treated with different concentrations of IFN-γ prior to the addition of 10 μm amyloid-β_1-42 (○) or 10 μm amyloid-β_42-1 (●). Values shown are the mean percentage cell survival from triplicate experiments repeated 3 times (n = 9), ± SD.

Figure 3

Pre-treatment with IFN-γ increases amyloid-β_1-42-induced caspase-3 activity. Levels of caspase-3 activity in cortical neurons pre-treated with control medium (○), or with 100 ng/ml IFN-γ (●) prior to the addition of varying concentrations of amyloid-β_1-42. Also shown is the caspase-3 activity from neurons pre-treated with control medium (□), or with 100 ng/ml IFN-γ (■) and incubated with varying concentrations of amyloid-β_42-1. Values shown are the mean percentage fluorescence units from quadruplicate experiments repeated twice (n = 8), ± SD. ** = Caspase 3 activity significantly greater than untreated neurons incubated with amyloid-β_1-42 (p < 0.05).

Figure 4

Pre-treatment with IFN-γ increases HuPrP82-146-induced caspase-3 activity. Levels of caspase-3 activity in cortical neurons pre-treated with control medium (○), or with 100 ng/ml IFN-γ (●) prior to the addition of varying concentrations of HuPrP82-146. Also shown is the caspase-3 activity from neurons pre-treated with control medium (□), or with 100 ng/ml IFN-γ (■) and incubated with varying concentrations of HuPrP82-146scrambled. Values shown are the mean percentage fluorescence units from quadruplicate experiments repeated twice (n = 8), ± SD. ** = Caspase 3 activity significantly greater than untreated neurons incubated with HuPrP82-146 (p < 0.05).

Figure 5

IFN-γ increases levels of cPLA₂ in SH-SY5Y neuroblastoma cells. Immunoblot showing levels of cPLA₂ and phospholipase C in lysates made from untreated cells and cells that had been treated with 100 pg/ml of IFN-γ for 3 hours. The fractions of the original samples that were added to membrane are shown.