views updated


Although research initiatives on normal aging of the human brain are still in their infancy, experimental studies have revealed a discrete spectrum of structural, physiological, and neuropsychological alterations that accompany the senescent changes that occur in other organ systems. While neuronal atrophy is the most consistent and pervasive change associated with the aging process, investigations have documented the presence of additional abnormalities in neuronal and non-neuronal (glial cell) morphological parameters, as well as enhanced vulnerability of specific neuronal and glial subtypes to cellular injury. Regional glial-cell activation is seen during brain aging, particularly within subcortical white-matter structures. During this phase of senescence, there is an increase in the size, number, and structural protein complement of astrocytes, a class of glial cells that support neuronal survival and function by secreting growth factors and providing additional molecular cues.

Neuronal loss initially occurs only within a few selected brain structures, such as the hippocampus. Degenerative changes also exhibit regional and cell-type-specific predilections. These progressive alterations include swelling of neuronal pre-synaptic (axonal) terminals, loss of axonal integrity, deterioration of neuronal post-synaptic (dendritic) arbors and synaptic loss, widespread accumulation of age-related pigments due to progressive oxidative damage, and the presence of large cytoplasmic inclusions, vacuoles, and granules containing polysaccharide moieties. Ultrastructural studies of these latter granular deposits have revealed the presence of fibrillar material resembling degenerated membrane-like structures that contain unbranched polysaccharide chains covalently linked to specific proteins.

Separate studies on human pathological materials and using mouse model systems have also shown that aging is associated with dramatic, regional profiles of dendritic growth and regression. These dynamic changes represent compensatory responses of healthy, surviving neurons to selective patterns of neuronal atrophy, biochemical alterations, and adjacent neuronal and glial cell loss that, collectively, impair normal cellular maintenance functions. Specific forms of synaptic plasticity (i.e., hippocampal long-term potentiation) are also compromised in the aging brain.

Experimental mouse strains have been utilized to examine the biological basis of brain aging, and such strains now serve as model systems to examine the potential applications of pharmacological- and gene-based therapeutic strategies. Specific inbred mouse strains, termed senescence-accelerated mice (SAM), exhibit apparently normal neural development, but later in life display the neuropathological profiles seen in human aging. These animal models provide the necessary experimental reagents to examine longitudinal changes, individual variations, and neurobehavioral correlates associated with the aging process. Both SAMP-8 and SAMP-10 mice display impairments in shock-motivated simple avoidance and in conditional-avoidance T-maze tasks, but there is an absence of clear deficits using more complex behavioral paradigms. Autoimmune mouse strains have been utilized to demonstrated the presence of age-related increases in brain-reactive antibodies that are also linked to deficits in avoidance-learning paradigms. These findings are consistent with an immunological model of aging, and may help to explain the presence of chronic inflammatory responses in pathological specimens from Alzheimer's disease patients.

The reactive increases in astrocyte size and cell numbers, as well as inflammatory changes, seen in the aging brain have been associated with increased levels of expression of the astrocyte markers S100β and glial fibrillary acidic protein (GFAP), and of the cytokine interleukin-6 (IL-6). Genetically-altered (transgenic) mice with IL-6 targeted to astrocytes exhibit neurodegenerative changes similar to those present during brain aging, whereas transgenic mice with IL-6 targeted to neurons, S100β targeted to astrocytes, or mice with genetic ablation of GFAP show neuropathological and/or behavioral changes unlike those found in the senescent brain.

Age-associated pathological changes in the brain are also thought to result from the cumulative effects of oxidative damage. Oxidative stress results in the formation of very reactive superoxide radicals. These unstable molecular species are catalyzed to hydrogen peroxide by the enzyme superoxide dismutase (SOD), and are eventually cleared by catalase or selenoglutathione peroxidase. Two competing theories suggest that elevations in SOD may either protect against oxidative damage and retard brain aging, or induce an imbalance in antioxidant enzymes and foster the enhanced production of hydroxyl radicals and ensuing oxidative damage. Various studies have suggested that levels of SOD1 and SOD2 may either be increased, decreased, or remain unchanged within the aging brain. Studies using SOD1 transgenic and gene deletion (knockout) mice suggest that SOD may be neuroprotective, and that deficiency of the enzyme may increase neuronal vulnerability to environmental stressors. The enzyme monoamine oxidase (MAO) normally produces hydrogen peroxide as a by-product of oxidative degradation of biogenic amines, which are central molecular effectors of neural communications. MAO-B levels are increased during brain aging and are thought to contribute to neurodegeneration through enhanced production of hydrogen peroxide. Selective atrophy of a specific subclass (dopamine-containing) neurons in a distinct brain region is, in fact, present in MAO-B transgenic mice. Elevated levels of corticosteroids are also found during brain senescence, and are linked to enhanced astrocyte reactivity, diminished elaboration of dendritic arbors, and ensuing neurodegeneration within the hippocampus. Growth-hormone transgenic mice exhibit accelerated brain aging as assessed by the finding of impairments of a variety of structural and neurochemical parameters.

There are also age-related deficits in a variety of neurobehavioral systems that are associated with linear declines in oxygen consumption in focal brain regions, such as the neocortex and the left thalamus. Prominent impairments in memory occur, but display significant subsystem dissociations.

Deficits in episodic, or working, memories, defined as autobiographical memories for context-specific events, are present at relatively early stages of aging, whereas alterations in semantic memories, representing general knowledge and other linguistic skills, become obvious only at later stages of senescence. Within the area of working memory, a central executive subsystem appears particularly vulnerable to age-associated impairments. Significant deficits in free recall are partially ameliorated with cued recall, however, and essentially disappear with the application of simple recognition tasks. Although older individuals display better visual than verbal memory functions, significant abnormalities in memories for faces, for abstract objects, and for the spatial location of objects also occurs. Procedural memories, however, which are required to perform skilled motor or cognitive acts, remain intact. With aging, reductions in visual acuity and contrast sensitivity are seen, mediated primarily by changes in central, rather than peripheral, connections. There are also progressive impairments in the ability to attend to objects in the visual field, particularly in the presence of visual distractors, indicating age-associated difficulties in ignoring irrelevant stimulus information.

Functional neuroimaging studies suggest that older and younger individuals utilize different brain regions and associated cognitive strategies to perform similar neurobehavioral tasks. This observation may reflect compensatory functional reorganization in older subjects, particularly in their utilization of alternate physiological interactions between the frontal lobes. In performing diverse visual processing tasks, older individuals display the requisite activation of ventral visual processing pathways, which mediate facial and object perception, and dorsal pathways, which mediate spatial perception. However, when compared to younger subjects, older individuals exhibit additional activation of the prefrontal cortices and the left temporal cortex, as well as stronger feedback influences from the frontal to the occipital lobes. With aging, there are reductions in performance of visual search tasks. In this paradigm, younger subjects show increased activation of the posterior visual cortex, whereas older individuals display enhanced bilateral activation of the prefrontal cortices. Older individuals also exhibit reductions in the accuracy of short-term facial recognition. In this task, both younger and older subjects show bilateral occipitotemporal and prefrontal cortical activation, whereas older individuals exhibit greater activation of the left dorsolateral prefrontal and occipitoparietal cortices, coupled with failure of hippocampal activation at short delay intervals.

The altered profiles of left prefrontal (semantic encoding) and parietal (short-term memory storage retrieval) cortical activation in older individuals appears to reflect the greater need for stimulus elaboration and an increased demand on storage capacity in the aging brain. By contrast, older subjects also show reductions in long-term facial recognition, representing a failure of stimulus encoding due to an inability to activate the prefrontal cortices and the hippocampus during stimulus encoding. Thus, for this task, older individuals display a failure to engage the appropriate cognitive processing networks, as well as an inability to recruit additional networks to compensate for these deficits. Interestingly, in episodic encoding and retrieval of word pairs, a cued recall task, older subjects display reductions in brain activation within the encoding network, but also display markedly enhanced right/left prefrontal cortical interactions and additional activation of the prefrontal cortex and other brain regions.

Aging is therefore associated with a spectrum of alterations in neuropsychological skills that reflect complex changes in the fidelity of the requisite cognitive-processing networks, differential impairments of specific complementary neural subsystems, and compensatory changes in regional patterns of activation and interregional communications. These alterations represent sophisticated biological adaptations designed to facilitate optimal neural network communications. At the cellular and molecular level, age-associated deficits in brain and cognitive functioning may thus reflect the cumulative structural and metabolic consequences of senescence, coupled with the more recent finding of a global loss of gene silencing. This novel molecular mechanism normally functions to sculpt patterns of gene expression in differentiated cells, including neurons. Progressive loss of gene silencing in association with aging may lead to alterations in the functional stability, and in the viability, of these essential cellular mediators of neural communications.

Mark Frederick Mehler

See also Brain, Memory; Neurochemistry; Neurodegenerative Diseases; Neuroendocrine System; Neurotransmitters; Plasticity.


Eustache, F.; Rioux, P.; Desgranges, B.; Marchal, G.; Petit-Taboue, M.-C.; Dary, M.; Lechevalier, B.; and Baron, J.-C. "Healthy Aging, Memory Subsystems, and Regional Cerebral Oxygen Consumption." Neuropsychologia 33 (1995): 867887.

Grady, C. L. "Brain Imaging and Age-Related Changes in Cognition." Experimental Gerontology 33 (1998): 661673.

Guarente, L., and Kenyon, C. "Genetic Pathways that Regulate Ageing in Model Organisms." Nature 408 (2000): 255262.

Ingram, D. K., and Jucker, M. "Developing Mouse Models of Aging: A Consideration of Strain Differences in Age-Related Behavioral and Neural Parameters." Neurobiology of Aging 20 (1999): 137145.

Jucker, M., and Ingram, D. K. "Murine Models of Brain Aging and Age-Related Neurodegenerative Diseases." Behavioural Brain Research 85 (1997): 125.

Shimada, A. "Age-Dependent Cerebral Atrophy and Cognitive Dysfunction in SAMP 10 Mice." Neurobiology of Aging 20 (1999): 125136.

Takeda, T. "Senescence-Accelerated Mouse (SAM): A Biogerontological Resource in Aging Research." Neurobiology of Aging 20 (1999): 105110.

Toussaint, O.; Remacle, J.; Clark, B. F. C.; Gonos, E. S.; Francheschi, C.; and Kirkwood, T. B. L. "Biology of Ageing" BioEssays 22 (2000): 954956.