07 Dec 2007

The study of brain aging gets a fresh treatment in a pair of studies that use vastly different approaches to understanding how the march of time makes it mark on living systems. One study, from Randy Buckner and colleagues at Harvard University, shows that as the brain ages, regions that form functional units become increasingly disconnected. Similar changes have been seen previously in people with early Alzheimer disease (see ARF related news story), but the new work finds the disconnection occurring even in elderly people who are healthy and show no sign of AD. The loss of coordinated network activity occurs together with changes in the brain’s white matter and cognitive decline. The paper appears in the December 6 issue of Neuron.

The second study, from Stuart Kim of Stanford University, Palo Alto, California, and Kevin Becker of the National Institute of Aging, Baltimore, Maryland, catalogs aging in mice by measuring global gene expression in 16 tissues at different times. Their results identify some common aspects of aging in all organisms, but also show that mice and humans may age quite differently at the molecular level. That paper was published November 30 in PLoS Genetics.

In the imaging study, first author Jessica Andrews-Hanna and colleagues looked at the activity in distributed brain networks in a group of 93 healthy adults between the ages of 18 and 93 using functional MRI. The investigators specifically measured activity in the default network, a set of widely distributed brain regions whose activity oscillates in sync in normal brain. They found that the concurrent fluctuations in the fMRI signal in the most far-flung anterior and posterior regions of the default network were severely disrupted in their aged participants, and that the extent of disruption increased with age. Another characteristic of the default network, that of deactivation during a task requiring focused attention, was also attenuated with age. The researchers observed changes not only in the default network, but also in a different distributed network, i.e., the dorsal attention system. A third network, that of the visual system, was unaffected by age. The results suggest that some, but not all, brain networks are subject to breakdown with aging.

Work over the past several years, notably from Michael Greicius and colleagues, has established disruptions of the default network as one of the earliest signs of Alzheimer disease (see ARF related news story and ARF related conference story). The present study establishes that the breakdown occurs even in adults with no apparent sign of AD. A subset of seven of the elderly subjects were determined to have no amyloid pathology based on negative PET scans using the amyloid imaging marker Pittsburg compound B. Nonetheless, those subjects showed a distribution of network connectivity values similar to the larger aged group, suggesting that the decline occurs independently of AD pathology.

The results support the “disconnection hypothesis” of aging, which holds that cognitive decline with age results from progressively weaker contacts between brain regions that normally function together. This could occur because of white matter loss or demyelination in the axon tracts that connect the regions. The researchers used diffusion tensor imaging (DTI) to look at white matter structure, and these scans revealed that disruptions in default network function were indeed associated with changes in white matter. The scientists conclude that these changes could account for the loss of network connectivity, but concede that they cannot rule out other factors, such as changes in neurotransmitter levels, that could play a role as well.

The functional and structural findings correlated with the cognitive abilities of individual subjects measured over several domains, the investigators show. “These results suggest that cognitive decline in normal aging arises from functional disruption in the coordination of large-scale brain systems that support cognition,” they write. Such disruptions also occur in a number of diseases, including AD but also schizophrenia, ADHD, and depression. One avenue of future work will be to determine whether different diseases show distinctive patterns of loss of connectivity. In early stage AD, the authors note, loss of connectivity occurs preferentially in posterior regions linked to the medial temporal lobe. This differs slightly from the changes they observe with aging, which appear most robust in the anterior to posterior network.

For something entirely different, the second study features a detailed molecular study of aging in mice, courtesy of gene chip analysis of 8,932 genes in 16 different tissues in both male and female mice at four different ages. The effort, headed up by joint first authors Jacob Zahn and Suresh Poosala, produced a database of gene expression the investigators call AGEMAP (Atlas of Gene Expression in Mouse Aging Project).

Broadly, the results reveal that different tissues have distinct transcriptional profiles associated with aging: some tissues, such as thymus and eye, show dramatic age-related changes in gene expression, while others, such as liver, show relatively fewer changes. The cerebrum, hippocampus, and striatum all show relatively low numbers of genes that change with age compared to the spinal cord, cerebellum, and most other non-neuronal tissues. Analysis of tissue-specific expression patterns identified three general aging profiles, one shared by the spinal cord and cerebellum, which the researchers call the neural profile. The others involved highly vascular tissue and steroid-responsive tissues. The results indicate that aging proceeds by different pathways in different tissues, so that a mouse might experience aging not as one universal process, but as several processes going on in different tissues. At the same time, the rate of aging is coordinated between tissues, and the results indicate that, just as for humans, some mice age faster and some slower.

The next obvious question to ask is whether there are common pathways affected by aging, in either different tissues or even different organisms. A comparison of the mouse data to previous studies in other species came up with just one common gene set; it consisted of genes involved in the mitochondrial electron transport chain, the source of both energy and damaging free radicals in cells. “In this analysis, the ETC came up as the big winner,” Kim told ARF. “Genes in the ETC were downregulated in many different old tissues in mouse, humans, worms, and flies. If the animal was a 2-year-old mouse or an 80-year-old human, the result was the same, a twofold decrease in expression in that pathway.” That makes the ETC pathway a promising marker for assessing old age at the molecular level, he said.

A set of genes for lysosomal proteins showed a common trend to increase expression with age in humans, mice, and flies, though not worms. Like the mitochondria, the lysosomal system is a cellular module of great interest to researchers studying neurodegeneration. Its elevation could reflect an increased turnover of damaged protein in old age, the authors write. Becker and Mark Mattson (also at the NIA and a coauthor on the new study) last month reported an in-depth analysis of the gene expression data from central nervous system tissues (Xu et al., 2007).

The most controversial finding, according to Kim, is the team’s observation that the overall pattern of gene expression in aging mice bears little resemblance to the pattern observed in humans. For this study, the California group compared their data to three human studies, one of which was Bruce Yankner’s study on the aging human brain (see ARF related news story). The finding raises the thorny question of whether old mice, a common model for both human aging and age-related diseases, are a good proxy for old humans. The jury is still out on that question, Kim says. Because his result is a negative finding, he cannot draw a firm conclusion. The work is still in its early stages, he stresses, and it is possible that there are similarities that have been missed in this first analysis. To inspire further study, the AGEMAP data will be made freely available for additional analysis by other researchers.—Pat McCaffrey

New Technologies Help Define Old Age

Quick Links

Tools

Comments

Make a Comment

References

News Citations

Paper Citations

Further Reading

Papers

News

Primary Papers

Annotate