 |
 |
Here's part1 of 2 of an article which I had posted previously and screwed up! Part 2 will follow tomorrow. I think this is a very interesting article which raises some pertinent questions. Wendy T. ******************************* Magazine: The Journal Of NIH Research Title: Alzheimer's Disease As A Model Of Molecular Gerontology Author: Allen D. Roses Issue: April 1995 Alzheimer's Disease As A Model Of Molecular Gerontology Allen D. Roses* Abstract: APOE4, a naturally occurring allele of the gene encoding apolipoprotein E (ApoE), increases the risk and lowers the age of developing late-onset Alzheimer's disease (AD). ApoE is the major apolipoprotein in the nervous system, and it may be necessary for the growth, maintenance, and repair of axonal membranes and myelin, the fatty sheath that surrounds axons. ApoE occurs in many cell types outside the central nervous system (CNS). Within the CNS, ApoE is expressed in astrocytes and is also located in smaller amounts in the cytoplasm of neurons in the cerebral cortex. My laboratory group has postulated that the E4 variant of ApoE may interact abnormally with neuronal cytoskeletal proteins such as tau, favoring microtubule degradation and the formation of neurofibrillary tangles, which occur in the brains of people with AD. In this commentary, I suggest that naturally occurring polymorphisms of genes other than APOE not only contribute to genetic diversity, but also may contribute to adult-onset diseases other than AD. In 1993, we reported that APOE4, a naturally occurring allele of the gene that encodes apolipoprotein E (ApoE), increases the susceptibility for developing late-onset Alzheimer's disease (AD)1-6. Since then, many other groups of investigators have confirmed the finding7-13. The impact of the APOE4 allele has become more obvious as people who are homozygous (APOE4/4) or heterozygous (APOE3/4) grow older and develop disease at a significantly faster rate than their counterparts who carry the APOE2 allele or are homozygous for APOE3 (APOE2/3, APOE2/4, or APOE3/3). Now, it is useful to stop and reflect on an interesting and obvious corollary. A polymorphic difference in a common gene, APOE, has a large influence on susceptibility to AD, a common late-onset disease3,4. APOE4 is not a mutant gene that causes AD. Instead, the APOE4 allele increases the risk and lowers the age of onset of developing AD--with APOE4 homozygotes at higher risk than heterozygotes. In contrast, the APOE2 allele appears to decrease the risk and increase the age of onset for AD. Although there are examples of statistically defined susceptibility factors in other diseases such as heart disease and cancer, the large genetic effect of the APOE4 allele on the increased rate of development of AD may forecast a common theme: Polymorphisms of other common genes may regulate the rate of disease expression over time in other late-onset neurological and nonneurological diseases. During the past decade, we have used many naturally occurring DNA polymorphisms as research tools to identify and locate genes associated with particular diseases and to map the human, mouse, and other genomes. Yet the biological effects of most polymorphisms have been virtually ignored as factors important for the phenotypic diversity of a species except when they are mutations for adisease. We tend to use the somewhat derogatory term "housekeeping genes" for those genes that are expressed in many tissues (constituitively) and apparently are necessary for sustaining all cell types--as if these genes are less important than other, more interesting, organ-specific genes (sometimes called "luxury genes"). Many widely expressed genes have polymorphic forms. APOE, the gene for Huntington's disease, and many other genes are expressed in most tissues and may play a key role in the rate of development of late-onset diseases14-16. In essence, the polymorphisms at the APOE locus on chromosome 19 determine the rate of disease development and the age of clinical onset of AD. I believe there are two important lessons to be learned independent of the specific role of ApoE in AD. First, different protein isoforms expressed from polymorphic alleles may have slightly altered function, with cumulative effects as people age. This alteration could lead to variable rates of disease onset. It is well accepted that during embryogenesis, many genes are transiently expressed and are involved in different stages of the maturation of the organism. Understanding the regulation of gene expression has, therefore, become an important problem in developmental biology. However, adults continue to mature because of other mechanisms that generate genetic diversity. We are now on the threshold of dissecting not only the genetic diversity that contributes to adult-onset diseases, but also of identifying factors that may underlie biological differences among interactions between humans and their environment. The relationship between APOE polymorphisms and AD was less important when people did not live more than 70 years and AD was a less prevalent disease. But now that we have extended our lifespans, we have also uncovered new limitations that previously were not subjected to evolutionary pressures or natural selection. The same APOE4 allele that increases our age-dependent risk of AD must also convey some selective advantage over the other variants of APOE to maintain itself in more primitive populations, including the western Australian aboriginees, Fore tribesmen of New Guinea, and Yanomami Indians of northwestern Brazil17,18. The second important observation affects basic neurobiology and emerges from the sudden focus of attention on ApoE as a factor in brain metabolism of AD. There is no evidence that ApoE is expressed in brain neurons and, therefore, the conclusion is that ApoE is not involved in intraneuronal metabolism. This belief generated immediate and repeated criticism of our suggestion in 1994 that ApoE3 and ApoE4 interact differently with microtubule-associated proteins (MAP2 or tau) in neurons in the cerebral cortex. If ApoE does not get into the neuronal cytoplasm, the reasoning went, then interactions with the cytoskeletal proteins MAP2 or tau would be in vitro artifacts without biological significance. But our recent detection of ApoE in the cytoplasm of normal cortical neurons as well as its accepted high concentration in astrocytes creates questions about its normal metabolic function (discussed below) and raises an interesting and paradoxical scenario19. AD involves the brain, and the disease is characterized by the loss of synaptic density and cortical neurons. APOE, the relevant genetic factor, encodes a 34-kilodalton glycoprotein whose mRNA is not known to be expressed in neurons. Therefore, there must be specific mechanisms for neurons to take up extracellular or astrocyte-secreted ApoE. Different isoforms of a protein that are not made in neurons can nevertheless have a profound effect on the viability of neurons. Therefore, gene expression in neurons is not equivalent to disease expression involving neurons. This is a critical concept that is relevant to many current experiments that attempt to identify organ-specific candidate genes as the site of mutations that lead to disease of that organ. Many brain diseases have been assumed to involve genes uniquely expressed in the brain. In fact, in most cases, the mutated gene is widely expressed but its clinical effect on late-onset degeneration appears specific to certain subsets of neurons14-16,20-24. When does a normal genetic trait become a disease gene? A genetic trait can be defined as an inherited variable. A disease is a morbid constellation of symptoms and signs. Some traits are benign throughout life, but increased longevity may provide the opportunity for the effect of other traits to contribute to the development of a disease. We are now at the threshold of differentiating a disease process from a genetic variability of neuronal aging. Figure 1 illustrates the distribution of the age of onset of AD in 72 families with late-onset inherited disease and 198 control subjects (with sporadic AD) as a function of the five commonly inherited APOE genotypes5. (APOE2/2 represents less than 0.05 percent of the U.S. population, and there were not enough data to include it in the figure.) With the caveat that the APOE4 effect actually starts before 60 years of age12,13, Figure 1 represents the relationship between APOE polymorphisms and the age of onset of AD. The prevalence of AD in a population depends on the relative APOE allele frequencies, the longevity of the society, and perhaps ethnic and/or racial genetic differences in other susceptibility genes. Several points are readily apparent from these data. The mean age of onset for individuals in the population with the APOE4/4 genotype (approximately 2 percent of the population) is less than 70 years of age; for APOE2/3 individuals (approximately 10 percent of the population), it is more than 90 years of age. The mean age of onset for APOE3/4 individuals (approximately 20 percent of the population) is in the mid-70s. Thus, a difference in the APOE genotype alone stratifies the distribution of age of onset of AD, with a mean difference of more than 20 years. There is another fascinating point that can be appreciated from Figure 1. Even with the "best" genotype, APOE2/3, everyone would get AD if they lived to 140 years. So what is the relationship of APOE polymorphisms to "aging"? Other studies of the associations of APOE alleles with longevity have approached this question in reverse, but with the same general results. In a 1994 study of 338 French centenarians, Schachter et al.2 found that the allele frequency of APOE2 in the centenarians was double that of a non-aged French control group (0.068 to 0.1280) and that the allele frequency of APOE4 was about half (0.112 to 0.052). The relative proportions of APOE alleles remaining in the centenarians could be approximated by using Figure 1 to estimate the alleles left in the population at age 100. There are several possible interpretations for these data. The most obvious is that the process of AD occurs in everyone at a rate influenced by their APOE genotype. A genetic trait becomes a susceptibility gene for a disease when the individuals at risk survive long enough for the disease to be expressed. APOE4 increases risk and lowers the mean age of onset of AD. In contrast, APOE2 appears to protect against AD until well past current human lifespans. We are living longer today than at any previous time in human history. Genetic traits that are beneficial earlier in life may be selected for during the first 50 years but may generate difficulties in old age. Thus, time is the variable that has changed most significantly and that has led to the current epidemic of AD. APOE may be a particularly important genetic locus for dissecting age- or time-related metabolic processes that occur in late adult life; the APOE gene may be a window on molecular gerontology. Clinicians traditionally measure the duration of disease as dating from the apparent age of onset. Of course, we realize that the disease processes may start years before we observe the resulting morbidity. Using sophisticated laboratory and imaging techniques, we attempt to identify markers of disease prior to the actual appearance of clinical symptoms26. In fact, late-onset diseases may well start before development ceases. They may involve a lifelong process that manifests itself only when signs and symptoms develop. Late-onset AD is just such a disease. Two decades ago, when the expected age of survival was about 67 years in men and 72 years in women, AD was relatively uncommon. But with increased longevity, processes leading to disease that need more time to develop begin to appear. An isoform-specific effect of a protein (such as ApoE) on the distribution of the age of onset of a disease would go unnoticed without the effect of longer survival. In fact, in primitive populations where the allele frequency of APOE4 approaches the 0.4 level, which is frequently observed with AD, the individuals in the population do not live long enough to develop late-onset disease. For example, the allele frequency of APOE4 is 0.39 in western Australian aborigines, yet there have been only rare autopsies in aborigines over the age of 56, and none of these people had developed AD. These autopsies have been performed during the past 30 years at the Royal Perth Hospital Department of Neuropathology (Byron Kakulas, personal communication, 1994)18. Other diseases that demonstrate time-dependent onset may not develop at a linear rate, but may accelerate over time. For example, Stanley Prusiner and his colleagues have demonstrated that the relatively rare prion diseases--Creutzfeldt-Jakob disease (CJD), Kuru, and Gerstman-Schenker-Straussler disease--represent a degenerative neuropathy that may result from an accumulation of the beta-pleated-sheet form of the prion protein in affected brain regions. The lead time for the conversion of prion protein from its normal alpha-helix structure to the beta-pleated-sheet conformation may progress slowly for many years, and accelerate as more of the (beta}-pleated sheet form accumulates20,27. But the infectious nature of the prion diseases also illustrates a mechanism for reducing the time that it takes for the transition to occur: Transmission of prions can provide protein aggregates that serve as a seeding function for beta-sheet formation. Once symptoms develop, the disease course appears to accelerate rapidly, almost resembling a log function for the relationship of disease to time. Triplet-repeat diseases, such as Huntington's disease, dentato-rubropallidal-luyisan atrophy (DRPLA), and myotonic muscular dystrophy, also suggest a mechanism for regulating the age of onset and the apparent severity of disease14-16,20-24,28,29. Several genes are known to have variable numbers of trinucleotide repeats that, when larger than the normal range, can be associated with variable age of onset diseases. If the rate of disease development is proportional to the size of the DNA triplet repeats, then mildly affected patients who carry small numbers of repeats and develop initial symptoms late in life could never be followed long enough to develop the full range of disease that develops over decades. Larger numbers of triplet repeats catalyze the rate of disease so that the disease starts earlier and can proceed through a longer course of signs and symptoms. It is important to understand the role of ApoE and its normally occurring variants in neuronal metabolism. It is here that the study of the genetics has delineated a new area of neurobiology. From a practical point of view, the definition of AD as a universal result of aging could lead to an important therapeutic strategy. It may be possible to use drugs to mimic the protective mechanisms of the ApoE2 isoform and thus extend the distribution of age of onset of AD, which is equivalent to delaying the disease by 20+ years30. What might happen after age 100 is of little practical concern because only a small percentage of the population is likely to survive that long. The phenotypes of AD can be explained by APOE genotype and time. The literature frequently refers to the phenotypic manifestations of AD, beta-amyloid (A-beta) plaques and neurofibrillary tangles, as though they are relatively invariable. In fact, the brains of people with AD typically have great variations in their A-beta load. Nevertheless, some investigators propose that A-beta fibril formation is the central factor in the development of AD31,32. In the rare beta-PP717 mutation form of the disease, a transgenic mouse model for increased beta-PP (amyloid precursor protein) and A-beta production can lead to amyloid plaques and associated gliosis33. Currently, there is no evidence of behavioral abnormalities--dementia of AD--in these mice. Whether this mechanism operates in beta-PP717 AD is uncertain because there is no current evidence for increased A-beta deposition in patients with this mutation. Late-onset AD, without beta-PP mutations, may have a wholly distinct pathogenesis. We generally assume that the same plaques contain A-beta protein as well as ApoE. Many investigators would be surprised to find that immunostained serial sections from the cerebral cortices of AD patients may show both processes, although not necessarily in the same plaques34. In fact, it was during such studies that we noted ApoE immunoreactivity in cortical neurons in addition to plaques, blood vessels, and glia. We also found that the density of immunostained A-beta protein (amyloid load) is related to different APOE genotypes and to the duration of disease from clinical onset to death. This evidence suggests that the isoforms of ApoE affect the rate of deposition of amyloid and the apparent amyloid load at autopsy9,35. Adjacent sections of brain from patients with AD stained with multiple antibodies can illustrate interesting differences. For example, fully formed plaques have prominent A-beta immunoreactivity, but ApoE immunoreactivity may be observed in some plaques when A-beta staining is not observed. Without rehashing various theories of AD pathogenesis, suffice it to say that ApoE can occur in neuritic plaques early in their formation, whereas A-beta deposition typically becomes greater with time as a function of the type of ApoE that an individual produces: ApoE4 > ApoE3 > ApoE2. Thus, when we compare microscopic sections of brain from AD patients and from controls for their APOE genotype and the length of AD disease between onset and death, we can account for the large differences in A-beta load that others have reported. Although the A-beta plaque deposition in AD is overwhelming and impressive, its variability can be explained by the presence of specific ApoE isoforms. Therefore, the phenotype is consistent with and dependent on the genotype5. Similar mutant genotypes may lead to diverse disease phenotypes. ***************** (to be continued)