4 Divya D.A. Raj, Bart J.L. Eggen and Hendrikus W.G.M. Boddeke Department of Neuroscience, Section Medical Physiology, University of Groningen, University Medical Centre Groningen, Groningen, the Netherlands Gene expression studies have been instrumental in gaining insight in brain ageing. A recent study (Glorioso et al., 2011) highlighted the consistency in results across several molecular studies of human brain ageing using gene expression analysis. Despite different microarray chip platforms and brain areas, a remarkable conservation of age-regulated changes has been demonstrated (Lee et al., 2000; Lu et al., 2004). This suggests the presence of a specific, tightly controlled, age-regulated transcriptional gene expression programme. A recent paper by Colantuoni et al. (2011) showed that a wave of gene expression changes during foetal development decreases upon early postnatal life but is resumed upon ageing. Development-related pathways showed a significant overlap with age-regulated pathways, indicating that genes associated with developmental transcriptional programmes may have a dual role in promoting the ageing process in accordance to the ‘antagonistic pleiotropy’ theory. The authors speculated on how some developmental processes, such as synaptic pruning, mirror ageing phenotypes such as synapse loss and thus could extend from the same transcriptional programmes. The mechanisms regulating age-related changes in gene expression are largely unknown. Lu et al. (2004) found that DNA damage accumulates particularly in the promoters of genes that show low expression with age. Such affected promoter regions were found to be genes that play central roles in synaptic plasticity, vesicular transport and mitochondrial function and were thereby proposed to initiate a programme of brain ageing. The idea of a defined transcription programme that underlies ageing also implies that it might be regulated by controlled epigenetic mechanisms making it operate as a clock. Indeed, cytosine methylation used as an epigenetic marker (by genomic mapping for 5-hydroxymethylcytosine) identified loci that are systematically altered during neurodevelopment and ageing. DNA methylation patterns were shown to influence transcriptional states and cellular identity during both development and ageing (Szulwach et al., 2011). A global decrease of genomic DNA methylation with age has been reported (Numata et al., 2012). In addition to global hypomethylation, a number of specific loci such as p16INK4a are known to become hypermethylated during ageing. Epigenetic plasticity in DNA methylation–controlled processes influences activity-dependent gene regulation (Martinowich et al., 2003), learning and memory (Miller and Sweatt, 2007) in the central nervous system (CNS). Alterations in mechanisms of epigenetic programming can thus alter gene expression and induce functional behavioural changes in the ageing brain. The processes of epigenetic coding and orchestration of transcriptional changes during brain ageing are only beginning to be unravelled (Rando and Chang, 2012). Gene expression changes in the ageing brain thus most likely represent underpinnings that orchestrate the behavioural changes and cognitive decline that accompany brain ageing. Association of inflammation with ageing is generally acknowledged. It is interesting to note that several longevity polymorphisms associated with human ageing are mediators of inflammation. Various polymorphisms in inflammatory and immune response genes such as interleukin-6 (IL6), tumour necrosis factor alpha (TNFα), IFNγ, IL1β, TGFβ and C-reactive protein (CRP) have been shown to be associated with longevity in humans. A list of longevity candidates related to immune functions and their corresponding studies have been listed in Table 4.1. It would be very interesting to correlate the polymorphism data with functional expression of the protein involved in vivo to know the precise role of inflammation mediated by this protein in tissue ageing. Individuals producing low levels of pro-inflammatory cytokines (e.g. IL6 and IFNγ) or high levels of anti-inflammatory cytokines (e.g. IL10) have been associated with longevity, suggesting that an enhanced pro-inflammatory response is a risk factor for shortened lifespan (Jylhava and Hurme, 2010). Table 4.1 Inflammatory genes and human longevity: Polymorphism studies. This table summarizes the studies correlating human lifespan to polymorphisms in inflammation or immune-associated markers (references mention corresponding publications). The studies were carried out in centenarian groups from various populations. Where information is available, the possible function of the polymorphism is also mentioned. Note: Superscripted footnote numbers refer to numbered sources in the ‘References’ column. Chronic inflammation is also associated with increased risk for age-related cognitive decline and dementia in several species, including rodents (Gemma et al., 2005) and humans (Dik et al., 2005). Cognitive modalities affected by inflammation include learning (Hein et al., 2010) and memory formation and consolidation (Frank et al., 2010). Cytokines are known to act directly on neurons and affect their functions such as excitability and gene expression (Lisak et al., 2011). Pro-inflammatory cytokines particularly down-regulate the neuronal genes involved in learning and memory (Godbout et al., 2005). Prolonged inflammation and production of IL6 by astrocytes have been shown to suppress proliferation, survival and differentiation of neural progenitors (Vallieres et al., 2002). To date, the best documented link between neuroinflammatory cytokine production and cognitive impairment has been established for IL1β. Sustained IL1β overexpression in the hippocampus impairs contextual and spatial memory in mice (Hein et al., 2010). IL1β is also demonstrated to be the mediator of lipopolysaccharide (LPS) and chronic stress–induced cognitive dysfunction (Terrando et al., 2010). Also, increased hippocampal expression of IL1β results in age-induced impairment of long-term potentiation (Murray and Lynch, 1998). Caspase-1 is involved in pre-processing of mature IL1β. Transgenic mice lacking the expression of caspase-1 do not suffer loss of contextual memory upon ageing (Gemma et al., 2005). In aged rats, IL1RA, the receptor antagonist for IL1β receptor, prevents Escherichia coli–induced suppression of long-term memory (Frank et al., 2010). Other inflammatory mediators that have been most consistently linked to poor performance in individual cognitive capabilities, particularly in memory and executive functions, are CRP and IL6 (Schram et al., 2007). Neuroinflammation-induced cognitive impairment may promote late-life depression disorders as inflammatory markers such as IL6 and CRP are also associated with depression in the elderly (Stewart et al., 2009). Inflammatory markers thus show promising predictive potential for the onset of age-associated dementia (Ravalgia et al., 2007). Comparison of gene expression profiles from several species and independent studies showed that inflammation is a general hallmark feature of the ageing brain. To estimate the robustness of inflammation in the ageing brain, inflammation-associated genes were analysed in the expression profiles of aged mouse brain tissue. Of the up-regulated genes, a quarter could be assigned to an immune or inflammatory response in the neocortex and cerebellum. Transcriptional alterations varied only between the different brain regions. Up to three-quarters of the gene expression changes were at least partially prevented by calorie restriction. Remarkably, however, the effect of calorie restriction on age-associated alterations in gene expression was highly dependent on transcript class. Calorie restriction was largely shown to reverse changes in immune and stress response genes without affecting alterations in neuronal gene expression in the aged mouse brain (Lee et al., 2000). Such inhibition of brain inflammation and increased production of trophic factors might underlie retardation of the brain ageing process by calorie restriction (Lee et al., 2000). The prominence of inflammatory gene expression was also found in other gene expression studies addressing brain ageing of rats, mice and humans (Lee et al., 2000; Lu et al., 2004; Hickman et al., 2013). It is also notable that the age-related genes of neuronal origin are predominantly down-regulated whereas genes of glial origin are up-regulated during ageing. In this respect, it remains to be delineated if inflammatory components are glia derived in the ageing brain. It is possible that the process of brain ageing might affect neural and glial cellular compartments in different ways. It is also possible that ageing affects a particular cellular compartment of the tissue depending on the cell type–specific vulnerability, which could in turn cause responsive reactions in glial cells. It, however, remains to be elucidated whether inflammatory changes are intrinsic results of ageing in glial cells or responsive reactions to alterations in the post-mitotic neurons. In any case, it is plausible that a significant portion of the inflammatory gene expression in the ageing brain originates from the local innate immune cells of the brain. Microglia constitute the sentinel immune network of the CNS (Graeber and Streit, 1990). Until the past decade, the role of microglia in an immunologically silent environment was only anticipated. It is now known that microglial responses are diverse and depend on the context and nature of environmental stimuli. Microglia fulfil a variety of functions in the healthy adult CNS: patrolling the brain to detect pathogens, synaptic scanning to check neuronal health, influencing synaptic transmission and promoting adult neurogenesis. It is conceivable that the ageing brain requires considerable homeostatic support. More insight into the basic functions of microglia in the adult brain could aid in unravelling altered functions with ageing. In 2005, it was reported that the response of microglia to disruption of the blood–brain barrier (BBB) and brain injury could be mimicked by adenosine triphosphate. By using two-photon imaging, it was shown that microglia are highly active in the resting state and extremely vigilant of changes in the brain environment (Nimmerjahn et al., 2005). Upon detecting an abnormality in tissue homeostasis, surrounding microglia are activated and target their branches towards the site of injury. Shielding of damaged sites by microglia may serve a neuroprotective role, as shown in an ischaemic brain model where microglial protrusions form a barrier between healthy and injured tissue (Wake et al., 2009). The vigilant role that microglia play in a healthy adult CNS implies microglial function in continuous monitoring of the brain environment. Microglia can sense synaptic activity through their neurotransmitter receptors, and overt release of neurotransmitters can trigger microglia activation (Pocock and Kettenmann, 2007). Evidence from a facial nerve transection model suggested that microglia might be able to influence the adult neuronal network by altering synaptic transmission in a peripheral nerve transection model (Blinzinger and Kreutzberg, 1968). Activated microglia were shown to be involved in severing of afferent synaptic boutons from the surface of regenerating motor neurons, and this process was termed ‘synaptic stripping’. Evidence for synaptic stripping by cortical microglia was demonstrated in a focal cortical inflammation model system in which activated microglia closely apposed neuronal perikarya and apical dendrites and were found to displace approximately 45% of the axosomatic synapses. Although it has been argued that the role of microglia in synaptic stripping is a case of ‘guilt by association’, more recent evidence using in vivo two-photon imaging indicates functional proof of the interaction of microglia, with neuronal synapses showing that the dynamic nature of microglia is directed in monitoring synapses. These contacts were also shown to be activity dependent, being reduced in frequency by reductions in neuronal activity followed by the disappearance of the presynaptic bouton (Wake et al., 2009). High-resolution electron microscopy has also shown the participation of microglia in synaptic junctions along with astrocytes redefining the interaction as the ‘quad-partite junction’ (Schafer et al., 2012). Major histocompatibility complex (MHC) class I or Ib antigens are required to regulate synaptic pruning on neuronal bodies that undergo retrograde degeneration after axonal transection (Shatz et al., 2009). Synaptic stripping is now shown to be preceded by a decrease of synaptic activity. Furthermore, C1q, a complement component, has also been shown to mark synaptic boutons that require removal by microglia (Schafer et al., 2012). Several modalities of microglial interactions with synapses were also found to be altered by sensory experience (light deprivation and subsequent exposure) in the visual cortex of juvenile mice. This raises the intriguing possibility that microglia may contribute to fine-tuning the plastic capacities of individual synapses in the healthy brain in accordance to experience and thus being capable of modulating crucial brain functions such as learning and memory (Tremblay et al., 2010). Upon encountering immune stimuli, microglia perform an orchestrated and controlled activation programme. Such immune stimuli may originate in the brain or come from the periphery. Activation of microglia in the healthy brain leads to an orderly beneficial inflammatory response, unless prolonged, overt or directed to self-antigens as in autoimmunity (Graeber and Streit, 2010). Aged microglia show increased immunoreactivity to CD68, a lysosome-associated phagocytic receptor, Toll-like receptors (TLRs), MHC class II antigen, OX6, the matrix-remodelling enzyme matrix metalloproteinase-12, Cd11b and Cd11c integrins and cytokines in the brain (as reviewed in Luo et al., 2010). In the ageing brain, microglia have been proposed to acquire an over-reactive phenotype resulting in the exaggerated immune response called microglial ‘priming’. Mounting evidence in several models of infection, injury and neurodegeneration indicates that the aged brain maintains a chronically increased level of inflammation. The best characterized and reproduced system for microglial priming is the response to peripheral infection. During a bacterial or viral infection, the neuroimmune axis comprising microglia communicates extensively with the peripheral adaptive and innate immune systems to induce sickness behaviour (Dantzer et al., 2008). The onset of this behaviour is an adaptive response in the animal to restore homeostasis. Mimicking infection by LPS treatment showed higher inflammatory cytokine production in primary glial cultures established from the brain of aged animals compared to young adults (Xie et al., 2003). In vivo studies with intraperitoneal injection of LPS or E. coli caused a prolonged and exaggerated cytokine response resulting in altered sickness behaviour in aged (22–24 months) Balb/c mice compared to young adults (Godbout et al., 2005). Cytokine mRNA levels in microglia isolated ex vivo from aged mice showed increased expression of pro-inflammatory mediators such as IL1β, TNFα and IL6, suggesting that the increased neuroinflammatory gene expression in the ageing brain might originate from microglia (Sierra et al., 2007; Henry et al., 2009). It is for this reason that microglia from aged brain were suggested to be ‘primed’ or ‘sensitized’ for activation and may cause excessive bystander injury to the aged brain, preventing functional recovering in the event of injury or insult (Godbout and Johnson, 2006; Perry et al., 2007). Microglia priming has been demonstrated in relation to several secondary stimuli in an ageing background as compiled in Table 4.2. This has been shown, for example, in studies on 1-methyl- 4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) mouse models, used to study Parkinson’s disease (PD), where aged animals were used. In an acute MPTP–PD model, old C57BL/6 mice (9–12 months old) were found to be more sensitive to neurotoxicity than young mice (3 months old), with more severe loss of dopaminergic neurons. Neurotoxicity observed in MPTP mouse therefore was shown to be age dependent (Sugama et al., 2003). Intraperitoneal administration of MPTP to old C57BL6 mice (14–15 months old) led to a remarkable loss of dopaminergic neurons with a marked decrease in dopamine levels (Phinney et al., 2006). Kainic acid–induced neurodegeneration and glial reactivity were also found to be more prominent in aged mice (Benkovic et al., 2006). These examples and the other studies summarized in Table 4.2 clearly demonstrate that microglia activation might indeed be more exaggerated in an ageing background. This suggests diminished control of microglia activation during normal ageing. Table 4.2 Hyperactive inflammatory profile under different stimuli in the ageing brain. This table summarizes studies involving varied stimuli in rodent models to demonstrate immune activation in aged animals. Most models demonstrate immune hyperactivation in the aged brain.
Ageing and the Immune Response in the CNS
Gene expression in the ageing brain
Inflammation as a hallmark of the ageing brain
Gene name and symbol
Group studied
Experimental size and age group
Longevity correlation
Possible function of the polymorphism
References
Interleukin-6 (IL6)
Italian
700 individuals; 60 to 110 years of age (323 centenarians)
Homozygotes for the G allele decreases in centenarian males, but not in centenarian females.
Negative association between the GG genotype of IL6 single-nucleotide polymorphism (SNP) and longevity in Italian centenarians.2
The −174 G/C SNP determines the transcription rate of IL6 mRNA.3 Among males, homozygotes for the G allele have higher IL6 serum levels in comparison with carriers of the C allele.1 The IL6 −174 GG genotype has been associated with higher plasma levels.4
Interferon gamma (IFNγ)
Italian
174 Italian centenarians > 99 years old and 248 < 60-year-old control subjects
The +874T allele, known to be associated with low IFNγ production, was found less frequently in centenarian women than in centenarian men or in control women, whereas no significant differences were observed in the distribution of the two alleles between male or female controls. Allele frequencies in centenarian men were not found to be significantly different from those of male controls.
The plasma levels of IFNγ bear a genetic regulatory component in which a 12 CA-repeat microsatellite allele in the first exon associates with elevated production in vitro.
Finnish
285 people, aged 90 to 95 years
This polymorphism is in absolute correlation with the T allele at the +874 T/A SNP site, and this T allele has been observed to be less frequent in centenarian women than in the control group.5 The frequency of allele G was higher in the survivors (n = 114) than in the non-survivors (n = 171).
Tumour necrosis factor (TNF superfamily, member 2) (TNFα)
Mexican
71 healthy elders aged 80 to 96 years; 99 young people (from 21 to 54 years; mean age 35.2 years)
The TNF2 allele was increased in the elder group when compared to young controls.
Not known
Soto-Vega et al. (2005)
Major histocompatibility complex, class II (HLA-DRB1, HLA-DQA1, HLA-DQB1)
Japanese (Okinawa)
Polymorphisms in Okinawan centenarians were analysed.
The DRB1*1401 allele was significantly increased in the centenarians, while the DRB1*0101 and DRB1*1201 alleles were slightly decreased.
The DQA10101=0104 and DQA105 alleles were significantly increased in the centenarians.
The DQB105 and DQB103 alleles were significantly increased in the centenarians.
Not known
Akisaka et al. (1997)
Toll-like receptor 4 (TLR4)
Sicilian
The G allele of this polymorphism has been associated with longevity.
The Asp299Gly (+896 A/G) functional polymorphism alters the receptor structure which leads to attenuation of the TLR4-mediated inflammatory response.
Balistreri et al. (2004)
Arbour et al. (2000)
Heat shock 70 kDa protein 1B (HSPA1B)
Danish
426 participants of various ages
Female carriers of the GG genotype survive better than noncarriers.
Not known
Singh et al. (2006)
Complement factor H (CFH)
This allele has been correlated to increased mortality, particularly cardiovascular morbidity (Laine et al., 2007).
The Tyr402His SNP (+1277 T/C) in CFH creates a functional ‘proinflammatory’ variant (402His) with markedly reduced CRP-binding capacity which predisposes the carriers to unbalanced and excessive inflammatory reactions via insufficient complement downregulation.
Jylhava. (2009)
Laine et al. (2007)
Interleukin-10 (IL10)
Italian
190 centenarians (159 women and 31 men > 99 years old) and 260 control subjects (99 women and 161 men younger than 60 years)
The −1082G homozygous genotype was increased in centenarian men but not in centenarian women. No difference was found between centenarians and control subjects regarding the other two SNPs.
Associated with high IL10 production
Lio et al. (2002)
Interleukin-10 (IL10)
Italian
72 centenarian men, 102 centenarian women and controls (115 men and 112 women, aged 22–60 years)
The number of male centenarians homozygous for the −1082G genotype was significantly increased in comparison with younger control subjects. No significant differences were observed between women and controls.
Suggested to be associated with high IL10 production
Lio et al. (2003)
Interleukin-10 (IL10)
Japanese
500 Japanese persons (mean age: 56.7 years old, range: 19–100)
There was a significant association of −819 T/C with age.
Not known
Okayama et al. (2005)
Nuclear receptor subfamily 3, group C, member 1 (glucocorticoid receptor) (NR3C1)
Dutch
402 men with a mean age of 77.8 years
After a follow-up of 4 years, 73 (19%) of 381 noncarriers died, while none of the 21 ER22/23EK carriers had died. Carriers may have lower C-reactive protein levels.
vanRossum et al. (2004)
Transforming growth factor, beta 1 (Camurati–Engelmann disease) (TGFB1)
Italian
419 subjects from northern and central Italy, including 172 centenarians and 247 younger controls
Significant differences were found at the +915 site as far as the C allele and GC genotype were concerned, both of them being lower in centenarians than in young controls, but none of the other tested genetic variants was significantly different between centenarians and controls. Moreover, a particular haplotype combination (G −800, C −509, C 869 and C 915) was notably lower in centenarians than in younger individuals.
Not known
Carrieri et al. (2004)
Microglia
Surveillance and motility
Microglial role in synaptic transmission
Microglia in the ageing brain
Model system
Priming stimulus
Animal
Findings
References
Peripheral inflammation via infection
Intraperitoneal injection of Escherichia coli lipopolysaccharide (LPS) (0.33 mg/kg, serotype 0127:B8)
Rats aged 24 months
Exaggerated and prolonged sickness behaviour and weight loss in aged rats.
Barrientos et al. (2009)
E. coli cultures (ATCC 15746)
Rats aged 24 months
Hippocampal functions such as memory of context, contextual fear, place learning and long-term memory consolidation are impaired in aged rats.
Barrientos et al. (2006)
Intraperitoneal injection of E. coli LPS (0.33 mg/kg, serotype 0127:B8)
Male BALB/c mice aged 20–24 months
LPS-induced elevation in the brain inflammatory cytokines and oxidative stress were both exaggerated and prolonged compared with adults.
Godbout et al. (2005)
Intraperitoneal injection of E. coli LPS (0.33 mg/kg, serotype 0127:B8)
Male BALB/c mice aged 20–24 months
Increased interleukin-1β (IL1β)-positive microglial cells and increased inflammatory response in the hippocampus of old mice. Cognition was affected in LPS-induced aged mice.
Chen et al. (2008)
Intraperitoneal injection of E. coli LPS (0.33 mg/kg, serotype 0127:B8)
Male BALB/c mice aged 20–24 months
Depressive-like behaviour in aged LPS-treated mice was associated with a more pronounced induction of peripheral and brain indoleamine 2,3-dioxygenase and a markedly higher turnover rate of brain serotonin.
Godbout et al. (2008)
Intraperitoneal injection of E. coli LPS (0.33 mg/kg, serotype 0127:B8)
Male BALB/c mice aged 18–20 months
Higher induction of inflammatory IL1β and anti-inflammatory IL10, IL1β, IL10, Toll-like receptor-2 (TLR2) and indoleamine 2,3 dioxygenase (IDO) mRNA levels in microglia isolated from aged mice than from adults. Increased MHC class II expression in aged microglia.
Henry et al. (2009)
Intraperitoneal LPS (1 mg/kg, 3 h)
Mice aged 18 months
Aging microglia contained lipofuscin granules, decreased processes complexity, altered granularity and increased mRNA expression of both proinflammatory (TNFα, IL1b and IL6) and anti-inflammatory (IL10 and TGFβ1) cytokines.
Sierra et al. (2007)
Adjuvant arthritis (AA)
Intradermal injection of complete Freund’s adjuvant
Rats
Proinflammatory microglia phenotype expressing ED1 and IL1β and deficits in the formation of LTP in the hippocampus of middle-aged rats.
Liu et al. (2012)
Viral infection
Intracranial injection of HIV-1 gp120
Male BALB/c mice aged 20–24 months
Behavioural deficits induced by gp120 were greater in aged mice than in adults.
Abraham et al. (2008)
Facial nerve axotomy
The right facial nerve was transected at the stylomastoid foramen.
Male F344/BN F1 rats aged 22–25 months
Age does not affect the glial response to axotomy in the lesioned facial nucleus.
Hurley et al. (2003)
Focal cerebral ischaemia
Middle cerebral artery (MCA) occlusion
MaleC57BL/Icrfat mice 26–31 months old
Cerebral oedema after ischaemia induction was seen in aged mice.
Fotheringham et al. (2000)
Intracerebral haemorrhage
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
Get Clinical Tree app for offline access
