SEARCH

SEARCH BY CITATION

Keywords:

  • cytokines;
  • cytotoxicity;
  • glial cells;
  • inflammation;
  • senescence;
  • transforming growth factor-β1

Abstract

  1. Top of page
  2. Abstract
  3. Aging
  4. Inflammation and microglial cell function in aging
  5. Neurodegenerative processes mediated by microglia: changes associated to aging
  6. Microglial cell reactivity and disease
  7. Perspectives
  8. Recapitulation
  9. Acknowledgement
  10. References

J. Neurochem. (2010) 112, 1099–1114.

Abstract

Among multiple structural and functional brain changes, aging is accompanied by an increase of inflammatory signaling in the nervous system as well as a dysfunction of the immune system elsewhere. Although the long-held view that aging involves neurocognitive impairment is now dismissed, aging is a major risk factor for neurodegenerative diseases such as Alzheimer`s disease, Parkinson`s disease and Huntington’s disease, among others. There are many age-related changes affecting the brain, contributing both to certain declining in function and increased frailty, which could singly and collectively affect neuronal viability and vulnerability. Among those changes, both inflammatory responses in aged brains and the altered regulation of toll like receptors, which appears to be relevant for understanding susceptibility to neurodegenerative processes, are linked to pathogenic mechanisms of several diseases. Here, we review how aging and pro-inflammatory environment could modulate microglial phenotype and its reactivity and contribute to the genesis of neurodegenerative processes. Data support our idea that age-related microglial cell changes, by inducing cytotoxicity in contrast to neuroprotection, could contribute to the onset of neurodegenerative changes. This view can have important implications for the development of new therapeutic approaches.


Abbreviations used
AD

Alzheimer disease

APC

antigen-presenting cell

APP

amyloid β precursor protein

Aββ

amyloid peptide

COX

cyclooxygenase

Fas

apoptosis-signaling cell surface antigen (TNFR family) cell death receptor

IFNγ

interferon-gamma

IL-1β

interleukin-1β

LI

1 μg/mL LPS + 10 ng/mL IFNγ

LPS

lipopolysaccharide

mGluR

metabotropic glutamate receptor

MHC

major histocompatibility complex

NFκB

nuclear factor-κB

NO

nitric oxide

NSAIDs

non-steroidal anti-inflammatory drugs

PAMP

pathogen-associated molecules pattern

PD

Parkinson’s disease

PPAR

gamma peroxisome proliferator-activated receptor gamma

PRRs

pattern recognition receptors

ROS

reactive oxygen species

SR

scavenger receptors

TGFβ

transforming growth factor-β

Th

T-helper cells

TLR

Toll-like receptor

TNFα

tumor necrosis factor-α

Aging

  1. Top of page
  2. Abstract
  3. Aging
  4. Inflammation and microglial cell function in aging
  5. Neurodegenerative processes mediated by microglia: changes associated to aging
  6. Microglial cell reactivity and disease
  7. Perspectives
  8. Recapitulation
  9. Acknowledgement
  10. References

The fact we are living longer presents us with new challenges, because advanced age is often accompanied by chronic diseases. Genetics and accumulation of mutations play a role, but epigenetic regulation (changes in gene expression that occur in the absence of alterations at the DNA level) is also a participating mechanism (Fraga et al. 2005; Fraga and Esteller 2007). Thus, potentially damaging agents are produced by the organism itself as inevitable consequence of metabolism (reactive oxygen species, ROS), or derived from the exposure to various physical (UV), chemical or biological (virus, bacteria, etc.) agents.

Aging can affect multiple tissues and processes, resulting in highly complex changes of function. It is known that some responses of the immune system decline with age, which leads both to impaired ability to fight infections and cancer. On the other hand, other properties are exacerbated increasing the probability of suffering autoimmune problems. This combination highlights the fact that more than a general decay in function, there is a dysregulation of the immune system with aging (Yung and Julius 2008). There is controversy about age-dependent decline in the function of macrophages (Gomez et al. 2008). Some authors remark the apparent hyperactivity of the brain innate immune response with the decreased function of the systemic immune system (Lucin and Wyss-Coray 2009; Valente et al. 2009; Weiskopf et al. 2009). We think that evidence points out to a dysregulated neuroimmune response, which shows an increased production of cytokines and cytotoxicity, but appears to be inefficient to solve challenges that would demand activation of phagocytic activity and neuroprotective effects. Our understanding of the cellular and molecular modifications that underlie age-associated changes, in particular those underlying the decline in nervous system function, have increased significantly in recent years. In this review, we discuss the changes observed in the nervous system with aging and the probable biological mechanisms responsible for them, with emphasis in the contribution of the brain innate immune system, the microglia.

Normal brain aging

There are multiple physiological changes occurring during aging (Table 1). Aging cells, not only in the nervous system, but everywhere, show accumulation of DNA damage, oxidative stress (Dröge and Schipper 2007), shortening of telomeres and the activation of tumor suppressor genes (Vijg and Campisi 2008). Normal aging is characterized by a mild chronic inflammatory activity with increased cytokine levels in blood and the brain and imbalance between pro- and anti-inflammatory cytokines, with decreased secretion of anti-inflammatory cytokines, such as IL-10 (Ye and Johnson 2001) and increased CNS levels of early proinflammatory cytokines such as tumor necrosis factor-α (TNFα) and interleukin-1β (IL1β) (Lukiw 2004; Streit et al. 2004b) and systemic IL-6 (Ye and Johnson 1999, 2001; Godbout and Johnson 2004). There are conflicting reports regarding changes on the level of inflammatory cytokines depending on aging and the presence of neurodegenerative processes (discussed in Rojo et al. 2009). Individuals with very high plasma IL-6 levels could have a greater tendency to exhibit a poorer baseline cognitive function and progress more frequently to cognitive impairment (Weaver et al. 2002). However, IL-6 levels do not strongly correlate with the magnitude of cognitive impairment (Dik et al. 2005), suggesting that increased cytokine levels are not secondary to pathological changes.

Table 1.   Age-dependent molecular changes modifying glial cell reactivity
Phenotype-functional effectMolecular changesReferences
  1. 1RELA, 1-Rel-NFκB transcription factor inhibitor; BDNF, brain-derived nerve growth factor; cPLA2, cytosolic phospholipase A2.

Increased pro-inflammatory cytokines Increased microglial cell reactivity Increased cytotoxicityIL-1β, IL-1α, TNFα, IL-6Colangelo et al. 2002; Conde and Streit 2006b; Godbout and Johnson 2004; Harry et al. 2000; Lovell et al. 2001; Lue et al. 2001; Lukiw 2004; Sierra et al. 2007; Streit et al. 2004b; Ye and Johnson 1999, 2001
Increased pro-inflammatory signalingDAAX, cPLA2, NFkB, Fas, COX2Adler et al. 2007, 2008 ; Colangelo et al. 2002; Lukiw 2004
Changes on anti-inflammatory cytokinesTGFβ (increased) IL-10 (decreased)Bye et al. 2001; Harry et al. 2000; Sierra et al. 2007; Ye and Johnson 2001
Decreased anti-inflammatory signalingBDNF, 1RELA, MAD3AColangelo et al. 2002; Zahn et al. 2007
Decreased cell viabilityDecreased secretion of neurotrophic factorsColangelo et al. 2002; Zahn et al. 2007
Change of PRRs expression patternChange of TLRs and SRs expression and signalingFortun et al. 2001; Hickman et al. 2008; Letiembre et al. 2007; Yamamoto et al. 2002
Impairment of NO regulationNO-dependent neurotoxicityJesko et al. 2003; Liu et al. 2004; McCann 1997
Decreased energy production Increased oxidative stressImpairment of mitochondrial electron transport chainDröge and Schipper 2007; Kurz et al. 2008; Navarro et al. 2002; Xu et al. 2008; Zahn et al. 2007
Decreased tolerance to toxins Reduction oxidative damageDecreased peroxisome genesMasters and Crane 1995; Zahn et al. 2007
Decreased cell reparationActivation of tumor suppressor genesVijg and Campisi 2008
Decreased cell proliferationIncreased expression of cell cycle arrest genesZahn et al. 2007

Gene expression and aging

Study of aging-related gene expression patterns (AGEMAP) revealed that the timing of changes associated with aging shows some degree of coordination in several tissues including the nervous system (Zahn et al. 2007). Although there is a high variability among species, genes involved in the electron transport chain showed common age-dependent regulation. The mitochondrial electron transport chain is the primary source of free radicals in cells, which are associated with oxidative damage that appears to be increased with aging (Dröge and Schipper 2007; Table 1). Lysosomal genes also showed a common increasing trend in expression with age. Being involved in degradation of membrane-bound and extracellular proteins, their increased expression may reflect increased turnover of damaged cell surface proteins in old organisms. In mice and humans, the inflammatory response and cytokine gene expression are increased with aging, showing the establishment of a pro-inflammatory condition. In contrast, peroxisome genes show a decreasing expression with aging (Table 1), which could lead to a decreased tolerance toward toxins and the amelioration of oxidative damage in old age (Masters and Crane 1995). Finally, genes involved in cell cycle arrest show an increased expression with age in a regionally distinctive manner, even at the CNS (Xu et al. 2007), implying the existence of areas of decreased ability for cell proliferation in the elderly (Zahn et al. 2007).

Oxidative stress and aging

In addition to the modification at the gene expression level, several of the above-mentioned factors interact among them for inducing senescence. Accumulation of ROS appears to be the most important causative factor for aging. ROS progressively increase with age, especially at the CNS because of its high oxygen demand and high metabolic rate. Progressive increase of ROS associates with the formation of radicals by mitochondria in the course of energy metabolism (Sohal and Sohal 1991; Barja 1999), which results in a decreasing availability of ATP for reductive synthesis and increasing peroxidization damage. Severe DNA damage caused by ROS, when not resolved, causes persistent DNA damage response signaling, which leads to p53-mediated senescence (Rodier et al. 2009). On the other hand, oxidative stress also causes elevation of calcium in the intracellular compartment, leading to activation of several calcium-dependent kinase, cytoskeleton changes, and catabolic and eventually death pathways. In the absence of neurodegenerative diseases, these aging-associated changes leave brain circuits intact but susceptible to synaptic damage and neuron degeneration (Hof and Morrison 2004; Mattson and Magnus 2006). There is an increase in the probability of developing a neurodegenerative disorder, such as Parkinson’s disease (PD) or Alzheimer disease (AD) after the sixth decade of life which is closely described by an exponential function (Jorm and Jolley 1998).

Iron accumulation and brain aging

Iron progressively accumulates in the brain with age (Zecca et al. 2004), which is normally associated to changes in iron metabolism and/or homeostasis. Iron accumulation affects protein modification, misfolding and aggregation (Hashimoto et al. 1999; Uversky et al. 2001). Besides, it has an important effect on production of radical species and eventually oxidative stress (Uversky et al. 2001; Hirose et al. 2003; Zecca et al. 2004). So iron accumulation becomes itself an aging promoting mechanism. The iron accumulation is quite specific and involves the accumulation of iron-containing molecules in certain cells, particularly in brain regions that are preferentially targeted by neurodegenerative diseases such as AD (Connor et al. 1995) and PD (Zecca et al. 2001). Interestingly, certain forms of iron accumulation are observed exclusively in glia (Connor et al. 1990).

Defects both at the protein level and oxidative stress can cause neurodegeneration and lead to the formation of the intracellular inclusion bodies that are one of the postmortem hallmarks of many neurodegenerative diseases (Rogers et al. 2002a; Lavados et al. 2008). As life expectancy increases, we should expect to see a progressive increase in the occurrence of iron-related neurodegenerative diseases, causing a vast range of disorders of the CNS.

Morphological changes in aging

There are many age-related physical changes at the nervous system. There are changes in gross morphology, associated with decrease of brain weight in the order of 2 to 3% per decade after the age of 50. The decrease in brain weight accelerates in later years, so in individuals of 80 years or older; the brain weight is typically decreased by 10% compared with that of young adults (see Drachman 2006). The Baltimore longitudinal study of aging, in which 138 non-demented subjects of 65-85 years old were evaluated with regular Magnetic Resonance Imaging analysis (Driscoll et al. 2009), also revealed patterns of changes in regional brain volume. In fact, this analysis allowed the differentiation between normal aging and mild cognitive impairment, a condition in which the person has certain level of memory impairment but not severe enough to interfere with daily life (Petersen et al. 1999). Anatomapathological evaluation reveals that the normal old brain presents accumulation of senile plaques, neurofibrilary tangles, granulovacuolar degeneration, Hirano bodies, glial cell proliferation and neuropil shreds (Davis et al. 1999). At the cellular level, the number of neurons shows little change over the age range from 20 to 90 years, with less than 10% of neocortical neurons lost at the age of 90 (Pakkerberg et al. 2003). However, neuron volume can show a significant reduction (Stark et al. 2007), but never at the extremes observed in neurodegenerative diseases. Normal aging neurons from the hippocampus show little decrease in size. In contrast, in AD, nearly 60% of CA1 neurons are lost with marked atrophy (Gosche et al. 2002). Aging affects also the neuronal sub cellular level. Dendritic arbors and synaptic spines of cortical pyramidal neurons undergo age-related regressive changes in specific regions (Duan et al. 2003; Hof and Morrison 2004). Assessment of post-synaptic markers shows a 46% decrease in synaptic spine number and density in brains from individuals older than 50 years compared with those from younger (Jacobs et al. 1997). Being the mayor post-synaptic site of excitatory synapses, decreased spines reflect changes in synaptic density and function. Pre-synaptic markers also decline with age. In the rat hippocampus, expression of the pre-synaptic protein synaptophysine specifically decline with aging in perforant path inputs to the dentate gyrus. The magnitude of decline in synaptophysine correlates with deficits in spatial memory (Bondareff and Geinisman 1976; Geinisman et al. 1986). Also, enzymes that synthesize neurotransmitters such as dopamine, nor adrenaline and acetylcholine decrease with age (Carlsson 1987). Alterations of dendrites and synaptic contacts have been also described in patients with neurodegenerative disorders (Knobloch and Mansuy 2008).

Cognitive changes in aging

In terms of cognitive abilities, aging is commonly associated with a progressive decline and, eventually, with the development of dementia (Reuter-Lorenz and Lustig 2005). In special, the acquisition of complex learning paradigms are decreased, a process that starts as early as middle-age (Salat et al. 2005). However, cognitive decline is not a mandatory outcome; many elderly people keep their cognitive capabilities intact. The Baltimore longitudinal study of aging, a prospective study with more than 2000 participants followed for up to 10 years with neurological and neuropsychological examinations, showed a wide inter-individual variability in cognitive decline with age (Shock et al. 1984).

Oxidative stress and inflammation appear as causative factors for progressive aging-related decline in motor and memory functions (Forster et al. 1996; Navarro et al. 2002), at least in part associated to a decrease of the mitochondrial electron transfer complex with the resulting decrease in energetic metabolism (Navarro et al. 2002). Over-expression of IL-1β, among other cytokines, and probably glial cell activation and changes on gliotransmission contribute to the impairment of cognitive abilities (Holmes et al. 2003; Tarkowski et al. 2003).

There is robust evidence showing that inflammatory cytokines, radical species and other mediating factors strongly influence neuronal function. Normal aging is associated with mild changes at synaptic and metabolic levels, and is also accompanied with changes of inflammatory activation at various levels, increased oxidative stress and changes on neurotrophic activity (Table 1). On the next section, we will discuss the effect of those changes on microglial cell function.

Inflammation and microglial cell function in aging

  1. Top of page
  2. Abstract
  3. Aging
  4. Inflammation and microglial cell function in aging
  5. Neurodegenerative processes mediated by microglia: changes associated to aging
  6. Microglial cell reactivity and disease
  7. Perspectives
  8. Recapitulation
  9. Acknowledgement
  10. References

Inflammation classically has been characterized by rubor (redness), tumor (swelling), dolor (pain), calor (heat) and funtio lesa (loss of function) at the various tissues of the body (Giunta 2006). In the brain, inflammation is defined mainly by an increased microglial cell reactivity (Lovell et al. 2001) as well as chronic increased levels of circulating cytokines such as TNFα and IL1β, and, at the CNS, increased levels of IL-1β, TNFα and transforming growth factor-β (TGFβ, a modulatory cytokine), all of which increase with aging (Lukiw 2004; Streit et al. 2004b). Microglia seem to be a major drive for brain aging (Fig. 1). Impairment of autophagocytosis and lysosomal proteolysis result in the formation of lipofuscin granules in both neurons (most abundant) and microglia (Sierra et al. 2007; Eichhoff et al. 2008; Xu et al. 2008). Accumulation of lipofuscin appears to be linked with increased oxidative stress and mitochondrial dysfunction (Brunk and Terman 2002; Szweda et al. 2003; Kurz et al. 2008). In turn, mitochondria dysfunction reduces the rate of ATP production and increases ROS, which has led to propose ‘the mitochondrial–lysosomal axis theory of aging’ (Brunk and Terman 2002; discussed in Nakanishi and Wu 2009).

image

Figure 1.  Age-dependent changes of microglial cells. In the aged brain, there is an increase of the number, size and activation of microglial cells, with increased basal but reduced induced phagocytic and lysosomal activity, and increased production of both inflammatory cytokines and radical species. Those changes result in a shift of balance with increased neurotoxicity. Additional pathophysiological changes associated with cardiovascular risk factors and metabolic syndrome, further promote an inflammatory environment, potentially increasing deleterious microglial cell activation.

Download figure to PowerPoint

In addition, increased microglial activation (Ye and Johnson 1999; Sierra et al. 2007) and elevated levels of inflammatory cytokines further induce cytokines and ROS (Table 1; Fig. 1).

Age-related dysfunction of microglial cells can be a significant cause underlying age-related neurodegenerative diseases such as AD (von Bernhardi 2007). Microglia display several morphological changes which are pronounced with aging. Senescent dystrophic microglial cells show decreased arborization and beading of their processes (Streit et al. 2004b, 2009) and abnormal cytoplasmic structures (Streit et al. 2004b; Conde and Streit 2006b). Senescent like-microglia colocalize with neurodegenerating neuronal cells and show functional deterioration, including a high incidence of clumping, particularly in white matter (Perry et al. 1993). It is especially interesting that senescent dystrophic microglial cells precedes the appearance of tau neuropathology (Streit et al. 2009). Microglia also show progressive replicative senescence characterized by an increased microglial proliferation in response to nerve injury (Conde and Streit 2006a) and inflammatory hyper responsiveness as a function of age (Streit 2005). Up to date, there is little characterization of the changes in microglial age-related identity markers profile, although there is evidence of increased expression of major histocompatibility complex (MHC)II and ED1 by aging microglia (Perry et al. 1993).

Microglial cell functions

Microglial cells constitute about 10% of adult CNS cell population and represent the innate immune system of the brain. They generate an antigen-non-specific response to injury and diverse endogenous and exogenous stimuli. Microglial cells constitute the first barrier and immune sentinels, being distributed through brain parenchyma continuously for sensing their microenvironment and producing pro- and anti-inflammatory cytokines. Microglia predominate in gray matter, with the highest concentration in hippocampus and substantia nigra (McGeer et al. 1988; Lawson et al. 1990).

Microglial cell populations are heterogeneous within different regions of the brain (Carson et al. 2007). In general terms, two populations of microglia can be distinguished. One, which is short-lived and frequently replaced from circulating monocyte-macrophages sources, is concentrated in perivascular and some specific parenchyma regions. The second population, which is long-lived, is resident and abundant in all the CNS parenchyma (Kennedy and Abkowitz 1997; Vallières and Sawchenko 2003). Of special interest for us is the fact that freshly hematogenously derived microglia appear to be more effective phagocytes than microglia obtained from brain resident population (Malm et al. 2005; Floden and Combs 2006; Simard et al. 2006; Town et al. 2008). We speculate that macrophages aging as brain resident cells can be different from those obtained from circulating newly generated cells. We also propose that Aβ deposition in AD could be due to the age-related impairment of microglial cells’ function.

Microglial cells that are basally quiescent are characterized by a small cell body and ramified processes. When activated in response to many stimuli, they undergo morphological changes that include enlargement of the cell body and shortening of cellular processes. Besides, microglial cells response is characterized by phagocytosis, T-helper cell (Th)2-type induction and through secretion of IL4, IL10 and TGFβ. Th2-activated cells produce IL4, IL6, IL10 and IL13, cytokines that promote humoral immune responses and down-regulate Th1-mediated responses, inhibiting numerous macrophage inflammatory functions (Town et al. 2005). The activated state of microglial cells, far from a single phenotype, represent a continuum change from innate to adaptive activation with the expression of different cytokines and cytokine receptors that modulate T cell response (Town et al. 2005). In activated state, microglial cells up regulate the expression of different cell surface activation antigens, pattern recognition receptors (PRRs), produce cytokines and secrete short-lived potentially cytotoxic species such as nitric oxide (NO) and ROS (Meda et al. 2001). NO has been implicated in neuronal damage, and the nitric oxide hypothesis of brain aging (McCann 1997) propose that NO could be one of the major causes of damage accumulation in the aging brain. However, it is known that NO mediates both neurotoxic (formation of peroxynitrite in an oxidative environment) and neuroprotective mechanisms (induction of vasodilatation in hypoxic insults). Although age-related changes of the NO synthase system appears to vary significantly (Jesko et al. 2003; Liu et al. 2004), it seems that aging attenuates the response of NO to hypoxia (Cañuelo et al. 2007), effect that could contribute to a reduced tolerance to hypoxia. On the other hand, microglia are the main source of oxidation products in the aging mice brain (Hayashi et al. 2008) and oxidative damage appears to be particularly detrimental for microglia (Lopes et al. 2008). In fact, abundant damaged mitochondria, which can generate abundant ROS, accumulate in microglia during aging, probably associated to the autophagic and lysosomal impairment already discussed.

Microglial cell response to inflammatory activation

As we already mentioned, glial activation results in diverse functional effects including proliferation, up-regulation of active molecules, release of cytokines and growth factors, phagocytic transformation and production of NO and ROS. Whereas early stages of an inflammatory response can protect neurons (Wyss-Coray et al. 2002), chronic inflammation and the subsequent activation of microglia become detrimental (Prinz et al. 1999; Schubert et al. 2000; Hanisch 2002). Chronically primed microglia exhibit more rapid induction and an exaggerated pro-inflammatory cytokine release, enhancing for example sickness behavior induced by lipopolysaccharide (LPS), and suggesting that aging microglia are over-responsive (Godbout et al. 2005; Dilger and Johnson 2008; Sparkman and Johnson 2008).

The PRRs participate in host defense response and phagocytosis of pathogen-associated molecules pattern (PAMP), and therefore are crucial for innate immune response. However, these receptors are also associated with neurotoxic microglial activation. Other PAMPs are Toll-like receptors (TLRs) and scavenger receptors (SR). Activation of TLR2, TLR4 and TLR9 induce NO production (Ebert et al. 2005) and SRA, SRBI and CD36 participate in microglial production of ROS in response to Aβ fibrils (Coraci et al. 2002; Husemann et al. 2002). There is a significative up-regulation of several TLRs (TLR1, TLR2, TLR4, TLR5, TLR7) as well as CD14 expression in the brain during aging (Letiembre et al. 2007), and conspicuous changes in the expression profile of SRs (Hickman et al. 2008).

Expression and binding to PAMP receptors also result in the expression of cell surface activation molecules such as MHC molecules MHCI and MHCII, CD80, CD86 and CD40, and the secretion of cytokines TNFα, IL6, IL12, IL18 (Town et al. 2005). These changes induce the antigen presenting cell (APC) phenotype characteristic of adaptive immune cells response (Gordon 2003; Varin and Gordon 2009) and cytotoxic microglial cell phenotype (O’Keefe et al. 2002). The interaction of CD40 and its ligand CD154 induces APC phenotype and inhibits phagocytosis of Aβ (Townsend et al. 2005). Blockade of this interaction reduces T cell/microglia-mediated neuronal damage in the context of experimental autoimmune encephalitis (Howard et al. 1999). Blockade of CD154 also increases the anti-inflammatory cytokines TGFβ and IL10 in the brain, and decreases production of pro inflammatory cytokines interferon-gamma (IFNγ) and IL2 by activated T cells in culture obtained from a mice model of AD subjected to Aβ immunotherapy (Obregon et al. 2008). Modulation of APC phenotype could be relevant for increasing phagocytosis of pathogenic molecules while avoiding cytotoxic effects of pro-inflammatory cytokines secreted by microglial cells and Th1 T cells.

Microglial cell reactivity and the profile of neurotrophic and inflammatory factors in aging

As brain ages, there is a continuum increase of stress-related gene expression with increased pro-inflammatory and decreased anti-inflammatory gene signaling in hippocampus (Table 1). Among the up regulated genes can be found DAXX (FAS binding protein, apoptosis mediator), cytosolic phospholipase A2, nuclear factor-κB (NFκB), Fas [apoptosis-signaling cell surface antigen (TNFR family) cell death receptor], IL1β, IL1α and cyclooxygenase2 (COX2), and among the decreased anti-inflammatory genes are reported brain-derived nerve growth factor, the NFkB transcription factor inhibitors v-rel reticuloendotheliosis viral oncogen homolog (1-Rel) and IKappa B-alpha (MAD3A) (Colangelo et al. 2002) (Table 1). NFκB appears to be especially relevant. The binding site for NFκB transcription factor, which have roles in immunity, inflammation and cell death, is a robust candidate for age-dependent gene regulation (Adler et al. 2007). Genetic blockade of NFκB in chronologically aged mice is capable of reversing the gene expression program and tissue characteristics to those of a young mouse (Adler et al. 2008).

Microglial cells appear to be basally activated in the elderly (Fig. 2) as well as in AD patient brains (McGeer et al. 1987). Microglial cells obtained by cell sorting from aging mice (18 months of age) present lipofuscin granules, decreased processes complexity, altered granularity and increased mRNA levels of pro-inflammatory cytokines TNFα, IL1β and IL6 and anti-inflammatory cytokines IL10 and TGFβ1 compared to young mice (2 months). Furthermore, pro-inflammatory challenge with a single dose of 1 mg/kg LPS (an endotoxin from Gram negative bacteria), induced an additional increase of the basal expression of TNFα, IL1β, IL6, and IL10 but not TGFβ (Sierra et al. 2007). This additional increase on cytokines levels could impair microglial function leading to an aberrant handling of pathogenic molecules (Fiala et al. 2005) and pre-dispose microglial cells to become cytotoxic and induce neurodegenerative changes.

image

Figure 2.  Microglial cells-dependent mechanism leading to neurodegenerative processes. Microglia can drive neurotoxicity through two mechanisms. First, microglia can initiate neuron damage by producing neurotoxic factors in response to nervous system damage or activation by inflammatory stimuli. Second, neurotoxicity could depend on the impairment of their protective mechanisms, including phagocytosis of cell debris, pathogens or other exogenous instigating stimuli and modulation of their activation. When the ability to activate these protective mechanisms fails, o when microglia are unable to respond to those mechanism, microglia will initiate a neuronal death program and drive a progressive neurodegeneration.

Download figure to PowerPoint

To study the systemic inflammatory effect and to try to mimic the inflammatory aging profile, peripheral injections of pro-inflammatory molecules, such as LPS, have been used in animal models. Treatment with a single dose of 5 mg/kg LPS induced an elevation of TNFα at the protein and mRNA levels and microglial activation in substantia nigra, cortex and hippocampus, which lasted for 10 months after the challenge (Qin et al. 2007). Furthermore, LPS also increased other pro-inflammatory factors such as TNFα, monocyte chemotactic protein-1, IL1β and NFκB p65. An age-related alteration in the signal transduction of TLR4 under inflammatory conditions has been also reported (Fortun et al. 2001), as well as an increased induction of CD14 and TLR4 after LPS treatment (Yamamoto et al. 2002).

Microglial cells: changes on neuroprotection and neurotoxicity can be age-dependent

It is especially interesting that pro-inflammatory genes are increased during aging and that this effect is accentuated in AD (Colangelo et al. 2002; Lukiw 2004). TGFβ1 concentration increases with age (Harry et al. 2000; Bye et al. 2001), apparently associated to the modification of glial activation observed during aging. It is worth noting that age-related microglial dysfunction can result not only in increased inflammatory reactivity and cytotoxicity, but also in the impairment of the neuroprotective functions they serve (secretion of cytokines and neurotrophic factors or phagocytosis of debris and exogenous molecules).

We have reported that microglial reactivity to Aβ (von Bernhardi and Ramírez 2001) and their phagocytic activity (DeWitt et al. 1998) are attenuated by astrocytes. However, glial activation by Aβ is markedly amplified by their inflammatory priming and, under those conditions; attenuation by astrocytes is not observed (von Bernhardi and Eugenín 2004). The inflammatory priming of glial cells, rather than Aβ aggregates, could be the principal trigger for glial activation and inflammation in AD. In the brain, Aβ interacts with both microglia and astrocytes and under physiological conditions microglial cell reactivity against Aβ would be mild and short lasting, being attenuated by the activation of astrocytes. However, under pro-inflammatory conditions, attenuation by astrocytes could fail thus enhancing microglial activation and secretion of inflammatory cytokines. Those results suggest that impaired astrocytes regulation in an inflammatory environment, as that that can be observed in aged animals, could enhance microglial cell activation. Impaired astrocytes regulation in an inflammatory environment, in turn, can increase cytotoxicity, amyloid β precursor protein (APP) synthesis and Aβ aggregation, and impair uptake and degradation of Aβ (Rogers et al. 2002b), establishing a vicious circle (von Bernhardi 2007), favoring Aβ accumulation.

In young and adult animals of a mice model of AD selectively over-expressing TGFβ1 in astrocytes (hAPP/TGFβ1), an amyloidogenic role of TGFβ1, potentiating Aβ production by astrocytes and accelerating amyloid angiopathy was revealed. However, TGFβ1 also modulates the innate immune response and decreases amyloid deposits in brain parenchyma, apparently through an effect mediated by phagocytic microglial cells (Wyss-Coray et al. 2001; Lesne et al. 2003). In this context, it is interesting to note that the activation of the TGFβ1 pathway seems to be impaired in AD. While TGFβ protein levels are increased in plasma and in CSF (Rota et al. 2006; Motta et al. 2007) mRNA levels of its principal effector Smad3 are decreased in hippocampus (Colangelo et al. 2002) and phosphorylated Smad2 and Smad3 have been shown to be misallocated in cytoplasm granules of hippocampal cells instead to translocate into the cell nucleus (Lee et al. 2006; Ueberham et al. 2006). Therefore, it is possible that at least some of TGFβ signaling pathways are impaired in aging microglial cells, while other pathways remain unaffected. Such unbalance could favor Aβ deposition and the activation of other signaling pathways like ERK that is also activated by TGFβ receptor I and participates in the modulation of IFNγ-induced NO secretion (Tichauer et al. 2007). However, recent studies in Tg2576 mice crossed with a dominant-negative TGFβ receptor II, that blocks TGFβ1 Smad2/3 signaling, showed a marked attenuation of amyloid deposits in brain parenchyma (Town et al. 2008). This observation seems contradictory with an anti-amyloidogenic role for TGFβ1. However, those results can be explained by the fact that TGFβ Smad2/3 signaling appeared to be only blocked in peripheral macrophages but not in microglial cells.

Impairment of glial cell activation and modulation, either because of decreased anti-inflammatory modulation or increased activation secondary to cytokines stimulation, oxidative stress or changes of the expression pattern of PRRs, result in an abnormal glial cell reactivity (Fig. 2). In turn, abnormal reactivity of microglia has the potential to favor a cytotoxic response, promoting the development of neurodegenerative changes.

Neurodegenerative processes mediated by microglia: changes associated to aging

  1. Top of page
  2. Abstract
  3. Aging
  4. Inflammation and microglial cell function in aging
  5. Neurodegenerative processes mediated by microglia: changes associated to aging
  6. Microglial cell reactivity and disease
  7. Perspectives
  8. Recapitulation
  9. Acknowledgement
  10. References

Cytotoxicity in aging

In vivo assays in rhesus monkey microinjected with fibrillar Aβ in the cortex showed that Aβ-induced neuronal loss, tau phosphorylation and microglial proliferation were observed only in aged monkey, whereas the same concentration of Aβ was not toxic for young monkeys. These results suggest that Aβ neurotoxicity in vivo is a pathological response of the aging brain (Geula et al. 1998). Microglial cells could actively participate in the genesis of neuronal damage (Block et al. 2007) in neurodegenerative diseases. Microglial cells appear to be primed in senescence in such a way that several additional stimuli can result in a persistent pro-inflammatory response capable of inducing neuronal death. Lesions similar to senile plaques and neurofibrillary tangles observed in AD patients can be also observed in non-demented elders (Arriagada et al. 1992). Those cortical plaques, mostly diffuse, can be abundant and show a similar distribution pattern (Knopman et al. 2003) to lesions of AD brains. However, plaques of non-demented elders do not show glial cell and inflammatory reactivity. Furthermore, evolution toward the mature senile plaques observed in AD is associated with microglial cell activation (Sheng et al. 1997; Wegiel et al. 2000). On the other hand, microglia from cognitively normal elderly individuals show morphological abnormalities (Streit et al. 2004a) and secrete several of the cytokines secreted by microglia exposed to Aβ (Lue et al. 2001). Positron emission tomography reveals microglial activation preceding AD cytopathology (Zhu et al. 2004), even in patients in the early stages of minimal cognitive impairment and AD (Cagnin et al. 2001). Microglial cell activation has been confirmed by autopsy studies of AD patients (Lue et al. 1996; Rogers and Lue 2001).

The peripheral increase of pro-inflammatory mediators can also initiate the synthesis of cytokines in the brain and have severe consequences in the elderly and demented population (Laye et al. 1994; Lerner et al. 1997). In mice, single peripheral administration of 5 mg/kg LPS reduced the number of tyrosine hydroxylase positive neurons in the substantia nigra by 47% at 10 months after treatment (Qin et al. 2007). In vitro assays have shown that soluble molecules such as TNFα secreted by microglial cells when exposed to pro-inflammatory molecules such as Aβ can induce proapoptotic effect on hippocampal cultures (Combs et al. 2001; von Bernhardi and Eugenín 2004).

Effect of systemic inflammation in animal models of neurodegenerative diseases

In the APPswe mice model of AD, the systemic administration of 0.5 mg/kg LPS increases Aβ and APP immune reactivity in specific intracellular compartments in neurons from neocortex and hippocampus (Sheng et al. 2003). Also, in the mice model of Prion disease ME7, characterized by an atypical inflammatory profile with a marked expression of TGFβ1, the systemic challenge with 0.5 mg/kg LPS shows induction of an exacerbated neuroinflammatory response, with increased levels of IL1β, TNFα, IL6, inducible NO synthase; changes that associates with increased neuronal apoptosis (Cunningham et al. 2005) and impairment of burrowing and locomotor activity (Cunningham et al. 2009). It is noteworthy that inhibition of TGFβ1 activity induces severe cerebral inflammation, indicating that TGFβ plays an important role in the regulation of microglial response (Boche et al. 2006). In all these models, changes were correlated with an increased activation of microglial cells located in the same region.

Microglial cell activation and early hyperphosphorylation of tau

In the transgenic model of AD (3xTg-AD), the chronic administration of LPS (0.5 mg/kg twice a week for 6 weeks), induces tau hyperphosphorylation in hippocampal neurons, effect that correlates with an increase in CDK5 activity (Kitazawa et al. 2005). Tau hyperphosphorylation is also observed in animal models exposed to mercury intoxication, mainly through the activation of JNK pathways, leading to neuropathological changes in the cerebral cortex, which shows increased numbers of microglia and astrocytes, but not in the hippocampus (Fujimura et al. 2009). The proposed mechanisms involve the activation of inflammation-associated signaling pathways, such as JNK. Interesting enough, methyl mercury also induces oxidative stress and cytotoxicity in microglia (Garg and Chang 2006). IL-1β (Li et al. 2003) as well as IL-6 (Quintanilla et al. 2004), both activators of the mitogen-activated protein kinase-p38 signaling pathway, can also induce an increase in tau phosphorylation and tangle formation, which seem to be mediated partially by mitogen-activated protein kinase p38. IL-18 also favors tau hyperphosphorylation by enhancing Cdk5/p35 and GSK-3β kinases (Ojala et al. 2008). We can speculate that the inflammatory environment of aged brain, through activation of specific signaling pathways could favor kinase activity responsible for tau hyperphorphorylation.

Microglial cell reactivity and disease

  1. Top of page
  2. Abstract
  3. Aging
  4. Inflammation and microglial cell function in aging
  5. Neurodegenerative processes mediated by microglia: changes associated to aging
  6. Microglial cell reactivity and disease
  7. Perspectives
  8. Recapitulation
  9. Acknowledgement
  10. References

Inflammation and glial activation in neurodegenerative diseases: AD

Age is the major risk factor for AD (Katzman 1986). Epidemiological evidence (Dziedzic et al. 2003; Sala et al. 2003) as well as that obtained using experimental models of AD (Griffin and Mrak 2002; Melton et al. 2003) show that pro-inflammatory conditions promote the development and progression of AD (Akiyama et al. 2000; Lim et al. 2000; Neuroinflammation Working Group 2000). Non-steroidal anti-inflammatory drugs decrease the risk of AD or delay its onset (Stewart et al. 1997; Broe et al. 2000; Etminan et al. 2003). Ibuprofen decreases cytokine-stimulated Aβ production (Blasko et al. 2001) and reduces neuritic plaques and inflammation in transgenic mice over expressing APP (Lim et al. 2000).

In neurodegenerative diseases such as PD or AD, the main circuits affected by neuronal lost are also densely populated by microglial cells (Kim et al. 2000). AD patients, even at early stages, show increased plasma levels of cytokines IL12, IL16, IL18 and TGFβ1 (Motta et al. 2007). IL16 stimulates the production of IL1β, IL6, and TNFα, and could participate in ischemic neuronal cell death (Lipton 1999); IL18 induces Th1 cytokines and in synergy with IL12 induce the production of IFNγ by natural killer and microglial cells (Lebel-Binay et al. 2000). On the other hand, TGFβ shows a pivotal role in the control of transition between Th1 and the anti-inflammatory phenotype Th2 response (Town et al. 2005).

TGFβ1 shows a low concentration in normal brain tissue, whereas its expression by activated glial cells is increased in the injured or diseased brain (Zhu et al. 2000; Dhandapani et al. 2003). IL1β, TGFβ and inducible COX2 are elevated in cerebrospinal fluid (Rota et al. 2006) and the brain tissue of AD patients (Luterman et al. 2000; Ho et al. 2001). Astrocytes isolated from AD patients (Blasko et al. 2004) and reactive astrocytes from AD-transgenic mice secrete TGFβ1 and IL10 (Apelt and Schliebs 2001). On the other hand, reduced levels of Smad3 are observed in AD patients (Colangelo et al. 2002; Katsel et al. 2005), suggesting that high levels of TGFβ1 could be a compensatory mechanism. The induction of TGFβ1 by pro-inflammatory molecules would limit the temporal and spatial extent of the inflammatory response (Ramírez et al. 2005; Saud et al. 2005). However, there is also reports that TGFβ1 production can be inhibited by TNFα (Blasko et al. 2004), and it is possible that at least certain inflammatory environments reduces or neutralize its protective response. On the other hand, administration of IL1β increases TGFβ1 in the brain, having a synergic effect with the neuroprotective activity of nerve growth factor (Chalazonitis et al. 1992; Vincent et al. 1997). However, besides TGFβ1 mediated neuroprotection, there are also potential deleterious effects. TGFβ1 also stimulates APP synthesis by astrocytes (Rogers et al. 1999); an effect that could be especially relevant considering the high level of TGFβ1 observed in aged individuals and its signaling impairment in AD.

The regional and cellular differences in the cytotoxicity of microglial cells could depend on multiple factors related to cell identity and environmental conditions. One of the factors could be the differential secretion of TGFβ and the increased levels induced by injuries (Zhu et al. 2000, 2002). Interestingly, TGFβ1 has various functions as an endogenous neurotrophic factor (Mouri et al. 2002; Hayashida et al. 2004; Ishida et al. 2004). Studies performed by our group have shown that in vitro conditioned media from neonatal rat microglial cells exposed to 2 μM Aβ induced neuronal apoptosis in hippocampal cultures. In contrast, conditioned media from mixed glial culture (microglia and astrocytes) exposed to 2 μM Aβ, did not induce apoptosis on hippocampal cultures (von Bernhardi and Eugenín 2004). Similarly, hippocampal neuron-astrocytes co-cultures treated with pro-inflammatory molecules LPS plus IFNγ (LI) showed a neuroprotective effect mediated by astrocytes (Ramírez et al. 2005). Thus, TGFβ1 appears to be one of the modulating factors favoring a protective effect, as suggested previously by the observation that TGFβ1 abolishes dendrite retraction and neurotoxicity (Eyüpoglu et al. 2003). Under pro-inflammatory conditions, hippocampal cultures containing astrocytes produce higher levels of TGFβ1. In contrast, induction of cell death is observed when hippocampal cultures containing microglial cells are treated with LI (Ramírez et al. 2008). In such conditions, LI promotes hippocampal damage, while microglial cells produce NO and O2•−, suggesting that activated microglial cells could generate a deleterious environment (Saud et al. 2005). TGFβ1 produced by neurons modulates several glial functions including microglial cell activation (Chen et al. 2002; Mittaud et al. 2002; de Sampaio e Spohr et al. 2002; Eyüpoglu et al. 2003) and decreases their production of NO and O2•− (Herrera-Molina and von Bernhardi 2005). We have also reported that microglial cell modulation by astrocytes’ conditioned media is abolished when the media is neutralized with a TGFβ1-specific antibody (Herrera-Molina and von Bernhardi 2005) in accordance with previously reported roles of TGFβ1 on the inhibition of the production of proinflammatory molecules IL-1β and TNFα and NO release (McCartney-Francis and Wahl 2002; Lieb et al. 2003).

Anti-inflammatory studies for neurodegenerative diseases in animal models and human trials.

Epidemiological studies have shown that the risk of suffering from AD is reduced or at least delayed in individuals chronically treated with non-steroidal anti-inflammatory drugs (NSAIDs). Evidence in animal models has shown that COX2, the rate-limiting enzyme in the formation of prostaglandins has an important role in neurodegenerative process. COX2-deficient mice have shown to be less prone to develop 1-methyl 4phenyl-1,2,3,6-tetrahydropyridine (MPTPs) neurotoxicity of dopaminergic neurons from substantia nigra than wild type mice (LeWitt and Taylor 2008). In the AD mouse model APP/PS1, over expression of COX2 promotes cognitive deficiencies that are prevented by the NSAIDs celecoxib (Melnikova et al. 2006).

Clinical trials using anti-inflammatory drugs such as prednisone, hydroxycloroquine and selective COX2 inhibitors celecoxib and rofecoxib had no effect on neurodegenerative disease progression. Probably these drugs might be protective only if they are administrated during mid-life, but once the neurodegenerative process is established anti-inflammatory drugs would have no effect on disease (Blennow et al. 2006).

Perspectives

  1. Top of page
  2. Abstract
  3. Aging
  4. Inflammation and microglial cell function in aging
  5. Neurodegenerative processes mediated by microglia: changes associated to aging
  6. Microglial cell reactivity and disease
  7. Perspectives
  8. Recapitulation
  9. Acknowledgement
  10. References

Microglial cells as therapeutic targets

Other than symptomatic palliative treatments, there are only a handful of effective treatments that modify in certain extent the underlying disease process for neurodegenerative disorders, and there is no effective treatment for repairing or regenerating a damaged brain yet. In all these diseases, glial cells are central contributors, even when their roles are often neglected when studying disease mechanisms or proposing therapeutic approaches. The understanding of how glial dysfunction of pathology contributes to neuronal dysfunction and degeneration, and vice versa (von Bernhardi 2007; Lobsiger and Cleveland 2007) is central for the understanding of neurodegenerative diseases.

The role of microglia as therapeutic target in neurological diseases is still a matter of debate. Although microgliosis and astrocytosis occur in neurodegenerative diseases, it is still discussed whether there is a causal relation and if that is the case in which direction; although probably a majority of the hypothesis poses microglial cells as deleterious cytotoxic participants. The question about the relevance of microglial cells in the progression of neurodegenerative diseases have been challenged by studies that pharmacologically ablate them or in transgenic animals in which microglial cells are selectively ablated (Carmen et al. 2006; Lalancette-Hebert et al. 2007; Grathwohl et al. 2009). There is no clear evidence that inhibition or reduction of microglia is beneficial. However, there is evidence showing that inhibition of microglia can be harmful, as is observed for regeneration models (Keilhoff et al. 2007; Potter et al. 2009) and demyelinating lesions (Glezer et al. 2007). The beneficial or deleterious effect probably depends on the type of activation undergone by microglia and the stage of the disease process. Furthermore, therapies that aim to reestablish an adequate modulation of the immune response (see von Bernhardi 2009) in elderly individuals could benefit even from modest changes, without requiring the restoration of immune function to the levels of young individuals.

Regulation of microglial cell activity through modulation of their phenotype

Evidence demonstrate that microglia have beneficial and neuroprotective activation profiles, associated to protection, maintenance and even repair. Therefore, the ideal therapeutic approach should involve attenuation of deleterious activation, but preservation of their beneficial effects. Microglial cell-mediated immune response in aging seems to be shifted to an APC phenotype and Th1 cytokine profile. Meaning that phagocytotic activity and neuroprotective effects are reduced, whereas cytotoxicity is increased. Then, the modulation of microglial cell activation toward an innate immune and Th2 response promoting anti-inflammatory activity could be more effective. An effective modulation of microglial cell activation with decreased oxidative damage and promotion of neuronal protection may be the key for AD therapy (Shie et al. 2009). In animal models, the blockade of CD40L by gene disruption or by neutralizing antibodies have shown to decrease APC phenotype, decreasing pro-inflammatory cytokines and increasing anti-inflammatory cytokines in the brain (Townsend et al. 2005; Obregon et al. 2008). Although anti-inflammatory treatment has shown effects only for the prevention of AD, but has failed as therapy for AD patients, peroxisome proliferator-activated receptor gamma (PPARgamma) has recently gained increased attention for the treatment of AD because of its function as a molecular target for NSAIDs.

New therapeutic approaches are also under scrutiny. Aβ immunotherapy protocols have revealed that immunomodulation and potentiation of the phagocytic activity of microglia could reduce disease progression at least on the animal model (von Bernhardi 2009). Agonists and antagonists of different metabotropic glutamate receptor (mGluR) subtypes appear to ameliorate microglial cell neurotoxicity in Experimental Autoimmune Encephalitis animal models, suggesting that regulation of mGluRs on microglia can be a target for therapeutic intervention in multiple sclerosis (Pinteaux-Jones et al. 2008). mGluRs are a promising therapeutic approach for various neurodegenerative diseases (Byrnes et al. 2009).

Pathophysiology of CNS neurodegenerative disorders is multifactorial (including oxidative stress, mitochondrial breakdown, protein aggregation, endosomal stress and inflammation among other mechanisms), therapies capable of modulating multiple pathophysiological pathways may prove more effective than those directed at a single target. Furthermore, because different senescent changes appear to predominate in different individuals, it is highly probable that a tailored treatment oriented toward individual age-related changes and impairments will be needed. Strategies oriented to reduce problems associated with protein aggregation that may be needed for some individuals could be non-useful in individuals in which reduced ability to quench ROS predominates as disease mechanism. Nevertheless, prevention of risk factors associated to increased severity of age-related changes probably still is one of the best available strategies to avoid pathological aging. Not an easy task considering the difficulties of identifying the most critical areas of potential dysfunction.

Recapitulation

  1. Top of page
  2. Abstract
  3. Aging
  4. Inflammation and microglial cell function in aging
  5. Neurodegenerative processes mediated by microglia: changes associated to aging
  6. Microglial cell reactivity and disease
  7. Perspectives
  8. Recapitulation
  9. Acknowledgement
  10. References

Aging is a matter of staying alive long enough for anyone, and will result in characteristic molecular and cellular changes. However, only some individuals will develop a neurodegenerative disease along their life. Age-related increase in oxidative, metabolic, inflammatory activation or other type of homeostatic stress, resulting in the accumulation of damaged subcellular structures, DNA and proteins, are involved in disease mechanisms. Several of those processes are associated to microglial activation, although the particular vulnerability of neurons appears to depend at least partially in their post-mitotic nature and their high rate of oxidative metabolism. Furthermore, determinants of vulnerability of specific neurons could include their metabolic or structural demand, existence of potentially toxic metabolic intermediates, repertoire of signal transduction pathways and stress protection mechanisms. Vulnerability is manifested in AD, where neuronal dysfunction and neurodegeneration leads to a profound compromise of cognitive abilities. There is evidence that both age-associated cognitive impairments and AD reflects vulnerability of similar neuronal circuits. However, whereas neuronal death predominates in neurodegenerative diseases, aging is mostly associated to synaptic changes. Regardless their similarities, in contrast to the alternative that AD may be an integral part of the aging process, it appears that age-related changes only increase the vulnerability of the nervous system but additional neuropathologic injurious stimuli needed to trigger AD must be also present.

References

  1. Top of page
  2. Abstract
  3. Aging
  4. Inflammation and microglial cell function in aging
  5. Neurodegenerative processes mediated by microglia: changes associated to aging
  6. Microglial cell reactivity and disease
  7. Perspectives
  8. Recapitulation
  9. Acknowledgement
  10. References