Defects in cytokine-mediated neuroprotective glial responses to excitotoxic hippocampal injury in senescence-accelerated mouse

Abstract

Aging is a result of damage accumulation, and understanding of the mechanisms of aging requires exploration of the cellular and molecular systems functioning to control damage. Senescence-accelerated mouse prone 10 (SAMP10) has been established as an inbred strain exhibiting accelerated aging with an earlier onset of cognitive impairment due to neurodegeneration than the senescence-resistant control (SAMR1) strain. We hypothesized that tissue-protective responses of glial cells are impaired in SAMP10 mice. We injected kainic acid (KA) to induce hippocampal injury and studied how cytokines were upregulated on Day 3 using 3-month-old SAMP10 and SAMR1 mice. Following microarray-based screening for upregulated genes, we performed real-time RT-PCR and immunohistochemistry. Results indicated well-orchestrated cytokine-mediated glial interactions in the injured hippocampus of SAMR1 mice, in which microglia-derived interferon (IFN)-γ stimulated astrocytes via IFN-γ receptor and thereby induced expression of CXCL10 and macrophage inflammatory protein (MIP)-1α, and activated microglia produced granulocyte–macrophage colony-stimulating factor (GM-CSF) and osteopontin (OPN). OPN was the most strongly upregulated cytokine. CD44, an OPN receptor, was also strongly upregulated in the neuropil, especially on neurons and astrocytes. KA-induced hippocampal upregulation of these cytokines was strikingly reduced in SAMP10 mice compared to SAMR1 mice. On Day 30 after KA injection, SAMP10 but not SAMR1 mice exhibited hippocampal layer atrophy. Since the OPN–CD44 system is essential for neuroprotection and remodeling, these findings highlight the defects of SAMP10 mice in cytokine-mediated neuroprotective glia–neuron interactions, which may be associated with the mechanism underlying the vulnerability of SAMP10 mice to age-related neurodegeneration.

Introduction

Brain function is based on complex intercellular signaling networks composed not only of neurons linked through synapses but also glial syncytia through which Ca²⁺ waves can pass (Scemes and Giaume, 2006). The microglia are a third type of cell that may participate in this intercellular signaling network (Bessis et al., 2007, Hung et al., 2010). Microglia are the resident immune cells of the central nervous system (CNS) and play major roles in the CNS immune network. Microglia detect normal and pathological levels of activity in the brain with short-term and long-term effects on how signals are transmitted through neuronal networks (Moriguchi et al., 2003). Microglia can be quickly activated to release inflammatory mediators, engage in phagocytosis, and participate in adaptive immunity. Astrocytes as well as microglia are rapidly activated in response to high levels of neuronal activity. Astrocytes strongly communicate with microglia and play a critical role in the activation of microglia. There is a good deal of evidence suggesting that interaction occurs between astrocytes and microglia (Lynch, 2009). Cytokine expression by astrocytes appears to be involved in microglia activation (Graeber and Streit, 2010).

Kainic acid (KA), a glutamate analogue, induces hippocampal injury via its excitotoxic effects and causes proliferation and activation of astroglial and microglial cells (Jorgensen et al., 1993). Pyramidal cells in the cornu ammonis (CA) 3 sector are particularly vulnerable to excitotoxicity in mice and rats treated with systemic, intracerebroventricular, or intrahippocampal injection of KA (Ben-Ari, 1985, Jeon et al., 2008, Oprica et al., 2003, Tauck and Nadler, 1985). When CA3 predominant hippocampal damage is induced by KA injection, CA3 pyramidal cells undergo degeneration and cell death, which result in axonal degeneration of the afferents to the CA1 pyramidal cell dendrites (Jorgensen et al., 1990, Nadler et al., 1980). In CA1, neuronal cell bodies are preserved but most of the afferent terminals are at least transiently lost. Neural degeneration gives rise to a characteristic set of morphological transformations of microglia into bushy cells in the stratum radiatum (s.r.) of CA1 and stellate cells in the s.r. of CA3. Neural degeneration also induces hypertrophic changes in astrocytes. These cellular transformations together with the upregulation of a wide variety of intercellular signaling molecules are known to occur most strongly around 3 days after KA injection (Borges et al., 2004, Kim et al., 2002), although early events mediated by pro-inflammatory cytokines start several hours after KA injection (Oprica et al., 2003).

Much evidence suggests that aging is the result of many forms of damage accumulation. Understanding of the cellular and molecular basis of aging requires exploration of the mechanisms causing damage to accumulate and the systems controlling such damage (Kirkwood, 2005). In this context, the immune system plays critical roles in aging, and aging may be associated with altered immune responses in the brain. Steady-state levels of inflammatory cytokines are increased in the brain with advancing age (Sparkman and Johnson, 2008). Microglia may acquire a reactive phenotype during aging (Dilger and Johnson, 2008). Conversely, microglia may not function efficiently in older animals (Conde and Streit, 2006).

The senescence-accelerated mouse (SAM) is a well-characterized animal model of human aging (Sprott and Austad, 2006) and has been widely used to investigate a variety of disorders associated with advancing age (Takeda et al., 1991, Takeda et al., 1981). SAM consists of prone (SAMP) strains that exhibit an accelerated senescence-prone phenotype that includes shortened life span and systemic age-related degenerative diseases as well as resistant (SAMR) strains that are senescence-resistant and have been used as usual aging controls. The SAMP10 strain has been established as an experimental model for the study of brain aging. Aged SAMP10 mice are characterized by impairments in cognitive behavior associated with cerebral atrophy (Shimada, 1999, Shimada et al., 1994, Shimada et al., 1992, Shimada et al., 1993). Individual neurons in the cerebral cortex of aged SAMP10 mice exhibit decreases in soma size, dendritic retraction, and loss of dendritic spines, resulting in loss of synapses (Shimada et al., 2003, Shimada et al., 2006), consistent with changes that occur in the aging human brain (Dickstein et al., 2007). Age-related progressive neuronal DNA damage, neuronal ubiquitinated inclusions, decreased tissue proteasome activity, and elevation of pro-inflammatory cytokine gene expression have been reported in SAMP10 mice (Kumagai et al., 2007, Shimada et al., 2008, Shimada et al., 2002). These microscopically and biochemically identifiable types of degeneration of neurons begin to appear around the age of 8 months in SAMP10 mice. Other strains of mice with usual aging processes such as SAMR1 and C57BL/6 do not exhibit remarkable age-related neurodegeneration. Our recent study revealed that hippocampal microglia of young SAMP10 mice have degenerated cytoplasmic processes similar to those in aged SAMR1 mice and more frequent pathological changes such as beading and fragmentation of processes (Hasegawa-Ishii et al., 2010). These morphological abnormalities in microglia of SAMP10 mice appear at the age of 3 months and precede the onset of neuronal degeneration.

In the present study, we hypothesized that intercellular signaling networks comprised of neurons, astrocytes, and microglia are impaired in young SAMP10 mice, and that dysfunction in the CNS immune network that has neuroprotective functions against tissue injury underlies the accelerated brain aging observed in this model. We injected KA into SAMP10 and SAMR1 mice and screened for upregulated cytokine-related genes, and identified tissue and cellular sources of upregulated molecules. Findings were compared between SAMP10 and SAMR1 mice to highlight the potential abnormalities in cytokine responses to excitotoxic hippocampal injury in SAMP10 mice.