I’ve mentioned before that, for some few people, exercise doesn’t seem to have a benefit, and the benefits of exercise for fighting age-related cognitive decline may not apply to those carrying the Alzheimer’s gene.

New research suggests there is another gene variant that may impact on exercise’s effects. The new study follows on from earlier research that found that physical exercise during adolescence had more durable effects on object memory and BDNF levels than exercise during adulthood. In this study, 54 healthy but sedentary young adults (aged 18-36) were given an object recognition test before participating in either (a) a 4-week exercise program, with exercise on the final test day, (b) a 4-week exercise program, without exercise on the final test day, (c) a single bout of exercise on the final test day, or (d) remaining sedentary between test days.

Exercise both improved object recognition memory and reduced perceived stress — but only in one group: those who exercised for 4 weeks including the final day of testing. In other words, both regular exercise and recent exercise was needed to produce a memory benefit.

But there is one more factor — and this is where it gets really interesting — the benefit in this group didn’t happen for every member of the group. Only those carrying a specific genotype benefited from regular and recent exercise. This genotype has to do with the brain protein BDNF, which is involved in neurogenesis and synaptic plasticity, and which is increased by exercise. The BDNF gene comes in two flavors: Val and Met. Previous research has linked the less common Met variant to poorer memory and greater age-related cognitive decline.

In other words, it seems that the Met allele affects how much BDNF is released as a result of exercise, and this in turn affects cognitive benefits.

The object recognition test involved participants seeing a series of 50 images (previously selected as being highly recognizable and nameable), followed by a 15 minute filler task, before seeing 100 images (the previous 50 and 50 new images) and indicating which had been seen previously. The filler task involved surveys for state anxiety, perceived stress, and mood. On the first (pre-program) visit, a survey for trait anxiety was also completed.

Of the 54 participants, 31 carried two copies of the Val allele, and 23 had at least one Met allele (19 Val/Met; 4 Met/Met). The population frequency for carrying at least one Met allele is 50% for Asians, 30% in Caucasians, and 4% in African-Americans.

Although exercise decreased stress and increased positive mood, the cognitive benefits of exercise were not associated with mood or anxiety. Neither was genotype associated with mood or anxiety. However, some studies have found an association between depression and the Met variant, and this study is of course quite small.

A final note: this study is part of research looking at the benefits of exercise for children with ADHD. The findings suggest that genotyping would enable us to predict whether an individual — a child with ADHD or an older adult at risk of cognitive decline or impairment — would benefit from this treatment strategy.

Data from the very large and long-running Cognitive Function and Ageing Study, a U.K. study involving 13,004 older adults (65+), from which 329 brains are now available for analysis, has found that cognitive lifestyle score (CLS) had no effect on Alzheimer’s pathology. Characteristics typical of Alzheimer’s, such as plaques, neurofibrillary tangles, and hippocampal atrophy, were similar in all CLS groups.

However, while cognitive lifestyle may have no effect on the development of Alzheimer's pathology, that is not to say it has no effect on the brain. In men, an active cognitive lifestyle was associated with less microvascular disease. In particular, the high CLS group showed an 80% relative reduction in deep white matter lesions. These associations remained after taking into account cardiovascular risk factors and APOE status.

This association was not found in women. However, women in the high CLS group tended to have greater brain weight.

In both genders, high CLS was associated with greater neuronal density and cortical thickness in Brodmann area 9 in the prefrontal lobe (but not, interestingly, in the hippocampus).

Cognitive lifestyle score is produced from years of education, occupational complexity coded according to social class and socioeconomic grouping, and social engagement based on frequency of contact with relatives, neighbors, and social events.

The findings provide more support for the ‘cognitive reserve’ theory, and shed some light on the mechanism, which appears to be rather different than we imagined. It may be that the changes in the prefrontal lobe (that we expected to see in the hippocampus) are a sign that greater cognitive activity helps you develop compensatory networks, rather than building up established ones. This would be consistent with research suggesting that older adults who maintain their cognitive fitness do so by developing new strategies that involve different regions, compensating for failing regions.

A number of studies, principally involving rodents, have established that physical exercise stimulates the creation of new brain cells in the hippocampus. A recent study attempted to uncover more about the mechanism.

Using two drugs that work directly on muscles, producing the physical effects of exercise, the researchers compared the effects on the brain. One drug (Aicar) improves the fitness of even sedentary animals. The other drug increases the effects of exercise on animals that exercise, but has little effect on sedentary animals.

After a week of receiving one of the drugs, sedentary mice performed better on tests of memory and learning, and showed more new brain cells. These effects were significantly greater for those taking Aicar.

Because the drugs have very little ability to cross into the brain, this demonstrates that the neurogenesis results from exercise-type reactions in the muscles, not to brain responses to the drugs. Indeed, previous research has found that direct infusion of Aicar into the brain impaired learning and memory.

Aicar increases the muscles’ output of AMPK, an enzyme that affects cellular energy and metabolism. It’s speculated that some of this enzyme may enter the bloodstream and travel to the brain. Interestingly, as with neurogenesis, AMPK activity in muscles appears to decline with age. It may be that AMPK production could serve as a biomarker for neurogenesis, as well as being a target for improving neurogenesis.

These findings add weight to evidence for the value of aerobic exercise over other types of exercise (given that the mice exercise by running). However, I see that human research has found that resistance training (which is difficult to study in mice!) also increases AMPK activity.

Do note — if you are hopeful that drugs will relieve you of the need to exercise — that the benefits were not only smaller than those achieved from exercise, but also didn’t last. In those mice taking Aicar for a second week, their brains not only stopped deriving any benefit, but actually deteriorated.

Genetic analysis of 9,232 older adults (average age 67; range 56-84) has implicated four genes in how fast your hippocampus shrinks with age (rs7294919 at 12q24, rs17178006 at 12q14, rs6741949 at 2q24, rs7852872 at 9p33). The first of these (implicated in cell death) showed a particularly strong link to a reduced hippocampus volume — with average consequence being a hippocampus of the same size as that of a person 4-5 years older.

Faster atrophy in this crucial brain region would increase people’s risk of Alzheimer’s and cognitive decline, by reducing their cognitive reserve. Reduced hippocampal volume is also associated with schizophrenia, major depression, and some forms of epilepsy.

In addition to cell death, the genes linked to this faster atrophy are involved in oxidative stress, ubiquitination, diabetes, embryonic development and neuronal migration.

A younger cohort, of 7,794 normal and cognitively compromised people with an average age of 40, showed that these suspect gene variants were also linked to smaller hippocampus volume in this age group. A third cohort, comprised of 1,563 primarily older people, showed a significant association between the ASTN2 variant (linked to neuronal migration) and faster memory loss.

In another analysis, researchers looked at intracranial volume and brain volume in 8,175 elderly. While they found no genetic associations for brain volume (although there was one suggestive association), they did discover that intracranial volume (the space occupied by the fully developed brain within the skull — this remains unchanged with age, reflecting brain size at full maturity) was significantly associated with two gene variants (at loci rs4273712, on chromosome 6q22, and rs9915547, on 17q21). These associations were replicated in a different sample of 1,752 older adults. One of these genes is already known to play a unique evolutionary role in human development.

A meta-analysis of seven genome-wide association studies, involving 10,768 infants (average age 14.5 months), found two loci robustly associated with head circumference in infancy (rs7980687 on chromosome 12q24 and rs1042725 on chromosome 12q15). These loci have previously been associated with adult height, but these effects on infant head circumference were largely independent of height. A third variant (rs11655470 on chromosome 17q21 — note that this is the same chromosome implicated in the study of older adults) showed suggestive evidence of association with head circumference; this chromosome has also been implicated in Parkinson's disease and other neurodegenerative diseases.

Previous research has found an association between head size in infancy and later development of Alzheimer’s. It has been thought that this may have to do with cognitive reserve.

Interestingly, the analyses also revealed that a variant in a gene called HMGA2 (rs10784502 on 12q14.3) affected intelligence as well as brain size.

Why ‘Alzheimer’s gene’ increases Alzheimer’s risk

Investigation into the so-called ‘Alzheimer’s gene’ ApoE4 (those who carry two copies of this variant have roughly eight to 10 times the risk of getting Alzheimer’s disease) has found that ApoE4 causes an increase in cyclophilin A, which in turn causes a breakdown of the cells lining the blood vessels. Blood vessels become leaky, making it more likely that toxic substances will leak into the brain.

The study found that mice carrying the ApoE4 gene had five times as much cyclophilin A as normal, in cells crucial to maintaining the integrity of the blood-brain barrier. Blocking the action of cyclophilin A brought blood flow back to normal and reduced the leakage of toxic substances by 80%.

The finding is in keeping with the idea that vascular problems are at the heart of Alzheimer’s disease — although it should not be assumed from that, that other problems (such as amyloid-beta plaques and tau tangles) are not also important. However, one thing that does seem clear now is that there is not one single pathway to Alzheimer’s. This research suggests a possible treatment approach for those carrying this risky gene variant.

Note also that this gene variant is not only associated with Alzheimer’s risk, but also Down’s syndrome dementia, poor outcome following TBI, and age-related cognitive decline.

On which note, I’d like to point out recent findings from the long-running Nurses' Health Study, involving 16,514 older women (70-81), that suggest that effects of postmenopausal hormone therapy for cognition may depend on apolipoprotein E (APOE) status, with the fastest rate of decline being observed among HT users who carried the APOe4 variant (in general HT was associated with poorer cognitive performance).

It’s also interesting to note another recent finding: that intracranial volume modifies the effect of apoE4 and white matter lesions on dementia risk. The study, involving 104 demented and 135 nondemented 85-year-olds, found that smaller intracranial volume increased the risk of dementia, Alzheimer's disease, and vascular dementia in participants with white matter lesions. However, white matter lesions were not associated with increased dementia risk in those with the largest intracranial volume. But intracranial volume did not modify dementia risk in those with the apoE4 gene.

More genes involved in Alzheimer’s

More genome-wide association studies of Alzheimer's disease have now identified variants in BIN1, CLU, CR1 and PICALM genes that increase Alzheimer’s risk, although it is not yet known how these gene variants affect risk (the present study ruled out effects on the two biomarkers, amyloid-beta 42 and phosphorylated tau).

Same genes linked to early- and late-onset Alzheimer's

Traditionally, we’ve made a distinction between early-onset Alzheimer's disease, which is thought to be inherited, and the more common late-onset Alzheimer’s. New findings, however, suggest we should re-think that distinction. While the genetic case for early-onset might seem to be stronger, sporadic (non-familial) cases do occur, and familial cases occur with late-onset.

New DNA sequencing techniques applied to the APP (amyloid precursor protein) gene, and the PSEN1 and PSEN2 (presenilin) genes (the three genes linked to early-onset Alzheimer's) has found that rare variants in these genes are more common in families where four or more members were affected with late-onset Alzheimer’s, compared to normal individuals. Additionally, mutations in the MAPT (microtubule associated protein tau) gene and GRN (progranulin) gene (both linked to frontotemporal dementia) were also found in some Alzheimer's patients, suggesting they had been incorrectly diagnosed as having Alzheimer's disease when they instead had frontotemporal dementia.

Of the 439 patients in which at least four individuals per family had been diagnosed with Alzheimer's disease, rare variants in the 3 Alzheimer's-related genes were found in 60 (13.7%) of them. While not all of these variants are known to be pathogenic, the frequency of mutations in these genes is significantly higher than it is in the general population.

The researchers estimate that about 5% of those with late-onset Alzheimer's disease have changes in these genes. They suggest that, at least in some cases, the same causes may underlie both early- and late-onset disease. The difference being that those that develop it later have more protective factors.

Another gene identified in early-onset Alzheimer's

A study of the genes from 130 families suffering from early-onset Alzheimer's disease has found that 116 had mutations on genes already known to be involved (APP, PSEN1, PSEN2 — see below for some older reports on these genes), while five of the other 14 families all showed mutations on a new gene: SORL1.

I say ‘new gene’ because it hasn’t been implicated in early-onset Alzheimer’s before. However, it has been implicated in the more common late-onset Alzheimer’s, and last year a study reported that the gene was associated with differences in hippocampal volume in young, healthy adults.

The finding, then, provides more support for the idea that some cases of early-onset and late-onset Alzheimer’s have the same causes.

The SORL1 gene codes for a protein involved in the production of the beta-amyloid peptide, and the mutations seen in this study appear to cause an under-expression of SORL1, resulting in an increase in the production of the beta-amyloid peptide. Such mutations were not found in the 1500 ethnicity-matched controls.

Older news reports on these other early-onset genes (brought over from the old website):

New genetic cause of Alzheimer's disease

Amyloid protein originates when it is cut by enzymes from a larger precursor protein. In very rare cases, mutations appear in the amyloid precursor protein (APP), causing it to change shape and be cut differently. The amyloid protein that is formed now has different characteristics, causing it to begin to stick together and precipitate as amyloid plaques. A genetic study of Alzheimer's patients younger than 70 has found genetic variations in the promoter that increases the gene expression and thus the formation of the amyloid precursor protein. The higher the expression (up to 150% as in Down syndrome), the younger the patient (starting between 50 and 60 years of age). Thus, the amount of amyloid precursor protein is a genetic risk factor for Alzheimer's disease.

Theuns, J. et al. 2006. Promoter Mutations That Increase Amyloid Precursor-Protein Expression Are Associated with Alzheimer Disease. American Journal of Human Genetics, 78, 936-946.

http://www.eurekalert.org/pub_releases/2006-04/vfii-rda041906.php

Evidence that Alzheimer's protein switches on genes

Amyloid b-protein precursor (APP) is snipped apart by enzymes to produce three protein fragments. Two fragments remain outside the cell and one stays inside. When APP is produced in excessive quantities, one of the cleaved segments that remains outside the cell, called the amyloid b-peptides, clumps together to form amyloid plaques that kill brain cells and may lead to the development of Alzheimer’s disease. New research indicates that the short "tail" segment of APP that is trapped inside the cell might also contribute to Alzheimer’s disease, through a process called transcriptional activation - switching on genes within the cell. Researchers speculate that creation of amyloid plaque is a byproduct of a misregulation in normal APP processing.

[2866] Cao, X., & Südhof T. C.
(2001).  A Transcriptively Active Complex of APP with Fe65 and Histone Acetyltransferase Tip60.
Science. 293(5527), 115 - 120.

http://www.eurekalert.org/pub_releases/2001-07/aaft-eta070201.php

Inactivation of Alzheimer's genes in mice causes dementia and brain degeneration

Mutations in two related genes known as presenilins are the major cause of early onset, inherited forms of Alzheimer's disease, but how these mutations cause the disease has not been clear. Since presenilins are involved in the production of amyloid peptides (the major components of amyloid plaques), it was thought that such mutations might cause Alzheimer’s by increasing brain levels of amyloid peptides. Accordingly, much effort has gone into identifying compounds that could block presenilin function. Now, however, genetic engineering in mice has revealed that deletion of these genes causes memory loss and gradual death of nerve cells in the mouse brain, demonstrating that the protein products of these genes are essential for normal learning, memory and nerve cell survival.

Saura, C.A., Choi, S-Y., Beglopoulos, V., Malkani, S., Zhang, D., Shankaranarayana Rao, B.S., Chattarji, S., Kelleher, R.J.III, Kandel, E.R., Duff, K., Kirkwood, A. & Shen, J. 2004. Loss of Presenilin Function Causes Impairments of Memory and Synaptic Plasticity Followed by Age-Dependent Neurodegeneration. Neuron, 42 (1), 23-36.

http://www.eurekalert.org/pub_releases/2004-04/cp-ioa032904.php