One of the most significant publications in the history of Alzheimer’s Disease (AD) research—and in genomics research generally—was just published online by Jonsson and colleagues in Nature magazine (1). I base this lofty assessment on impressive advances of two general types made by this study: those specific to AD, and others of even greater and more general importance. To whet your appetite for a more full taste of the latter, consider that the Nature title, “A mutation in APP protects against Alzheimer’s disease and age-related cognitive decline,” is compelling, but greatly undersells the magnitude of the discovery within. The logically equivalent title of this blog post, on the other hand, highlights the underlying reason this discovery is so notable.

More about this shortly, but, for now, let’s back up and consider how this study contributes to our understanding of the biology of AD (if you're not really interested in AD, but want to understand the far-reaching consequences of this discovery, you can jump to the next section below). The study reveals the identity of a variant of amyloid precursor protein (APP) that is protective against both AD and general cognitive decline. This variant results from an amino acid substitution of a threonine (T) for alanine (A) at amino acid position 673 in the APP protein near a key processing site involved in the production of amyloid beta (the variant is known as APP A673T). Therefore, when considered together with previous studies, this study appears to validate beyond doubt the previously controversial amyloid beta cleavage and aggregation hypothesis.

This is the first such protective minor variant [f1] ever discovered, and it previously had gone undetected because it is so rare—present in less than 1% of Icelandic and Scandinavian populations, and even rarer in North Americans. Bear in mind that APP A673T might exert maximum protective effects on Icelanders (Icelandic genetics and environments are unusual; the potentially complex interactions of genes and environments highlights the importance of the broadly inclusive information collection and aggregation approach taken by the PGP (2)). Nevertheless, for reasons explained below, we should expect either validation of the protective effects of this genetic variant or the discovery of other similar ones in non-Icelandic populations.

For context, contrast APP with APOE, the best known gene associated with increased risk of AD. APOE e3, the primary or major variant of APOE, is protective against AD, while the minor variant APOE e4 increases risk. The average worldwide frequencies of e3 and e4 are about 78% and 15%, respectively (3). This is considered to be the more typical case of a minor variant increasing risk of disease [f2], and deleterious minor variants are the focus of most recent studies and gene hunters. Also in contrast to APOE, APP variations from the typical alanine at amino acid position 673 (673A) appear to be extremely rare among humans; the results of Jonsson and colleagues suggest that the frequency of this predominant variant might represent about 99% of all variants at this amino acid position.

About 2% of the population are APOE e4 homozygotes and they are reported to have a 10 to 30-fold increased risk of AD relative to people who carry no e4 variants , so discovering this about yourself or a loved one can be quite worrisome (f3). However, Jonsson and colleagues report that APP A673T protection against AD appears to extend even to those with APOE e4 genotypes. The protective effect of APP A673T appears dramatic in other ways, too. The number of carriers who have undergone routine cognitive testing is small (and therefore current analyses are subject to great uncertainty), but one analysis of cognitive decline performed in this study indicates at least a 10 year cognitive advantage of carriers versus non-carriers between ages 80 and 100 years old.

Large-scale genome sequencing is shifting the discovery paradigm
It is important to consider the approach used to make the initial finding in this groundbreaking discovery: whole genome sequencing of a large number (1,795) of people. The belief that whole genome sequencing has a good chance of succeeding where other approaches [f4] have failed has met with increasing recent skepticism by some scientists (4). The discovery of APP A673T is a clear vindication of genome sequencing as a tool for discovery of both potential causes and cures of disease and debility [f5], and it underscores the existence of an often unrecognized but extremely important class of genetic variation: functional minor variants that provide protection against common diseases.

Some of these variants are likely to be so rare that large-scale genome sequencing is the only efficient means of discovering them. As I mentioned above, APP variations at alanine 673 appear to be quite rare in human populations—and this rarity probably extends far beyond humans. I performed a quick analysis of the evolutionary conservation of this alanine across many species and found that it is conserved across all mammals, including dolphins. It also appears to be present in chickens, turkeys, anolis lizards, turtles, and at least some zebrafish, although it is not found in many other fish. Given this level of evolutionary conservation we can infer that amino acid substitutions at this position are strongly selected against in mammals and certain other organisms.  It is very interesting and might be a bit surprising that a rare minor variant has such a strong protective effect relative to the predominant variant—and it is also why this discovery is so important.

Returning to the title of this post, consider the flipside of this protective phenomenon: the highly conserved and most common variant of APP increases the risk of AD and cognitive decline. But how is this possible?

How can extremely common variants be deleterious?
It is commonly assumed that natural selection purges most deleterious variants from the overall gene pool. This has been mostly true throughout human history, but nowadays most people live to ages far in excess of those shaped by evolutionary forces over that time. Gerontology is the study of aging and senescence, and gerontologists have accepted for decades that natural selection exerts diminishing selective pressure on the genes of an organism transitioning from reproductive to post-reproductive age.

In 1952 Peter Medawar proposed that deleterious variants will arise and persist in a population if negative effects occur very late in life (5). In 1957 George C. Williams extended this general framework with a critical insight: variants that contribute essentially to high levels of reproductive fitness early in life, but that also carry late-life deleterious effects, will persist through positive selection—even if these later effects are catastrophic (6) (Williams’ theory is often called Antagonistic Pleiotropy). In summary, there are two general and non-exclusive ways that gene variants (even very common ones) might exert negative effects in the post-reproductive period of life: 1) the absence of selection against certain variants because their late-life negative effects have little or no impact on reproduction, and 2) positive selection for variants that contribute reproductive benefits early in life with the unfortunate side effect of negative consequences after reproduction has ceased.

Common diseases and common variants reconsidered
When properly framed in this way, it is easy to understand how the vast majority of people might carry variants of APP that predispose to AD and cognitive decline late in life, while a rare minor variant is protective. We don’t yet know whether or how the more common APP variant is beneficial earlier in life relative to the A673T variant, but it is clear that A673T carriers are overrepresented in later life by several fold relative to non-carriers. One possibility for the evolutionary conservation of alanine 673 is that it carries substantial early life and reproductive benefits, especially in historical or more natural environments. It will be interesting to compare the reproductive successes and other phenotypes of the people carrying different variants.

It is important to understand that the evolutionary framework constructed by Medawar, Williams, and others has wide-ranging implications for the genetic determinants of disease and debility in older people. Most people who develop serious disease in developed countries are older and a few related diseases lead to the majority of mortality (cardiovascular diseases and cancer together account for about half of all mortality in developed countries). For non-gerontologists this discovery of APP A673T by Jonsson and colleagues provides a new twist in the public scientific discussion on common diseases and common variants. We should expect that the protective effects provided to older people are unlikely to be limited to this variant or to Alzheimer’s disease, and as we sequence more genomes we’ll surely discover more such variants protective against other diseases and general debility. Each time we do, there will be a logical complement or corollary for common and even predominant variants: they are risk factors for disease.

Protective variants provide a direct route to therapies
the most important information any protective variant provides is to reveal a specific way biology mitigates an extremely serious health risk, and therapies can be designed to reproduce its critical function. Pathogenic variants aren’t nearly so helpful. In the case of APP, drugs that mimic the effect of A673T have been in development for several years already (7) but no such drug is yet on the market or even in late stage clinical trials, but promising candidates are in early trials. The reason these drugs are already in the pipeline is that decades of basic biology had already pointed to one promising way to treat AD; so whole genome sequencing isn’t solely responsible for the eventual development of these therapies. However, given the very large protective effect of A673T for both AD and general cognitive decline, the market for mimetic therapies might be much larger than previously estimated—and their individual and societal benefits would be profound.

FOOTNOTES:
[f1] A major variant is relatively common in a population and a minor variant is uncommon.
[f2] Typical examples of protective minor variants provide protection against pathogen infection, often through loss of function (see 's Invulnerability to Stomach Flu Is My Secret Superpower). APP A673T is a functional minor variant which has different effects on a carrier’s intrinsic biochemistry and physiology relative to the major variants, so should be considered differently than variants selected for pathogen resistance.
[f3] Some reports suggest that the influence of APOE e4 on AD might be largely mitigated by routine exercise (8)
[f4] For example, genome-wide association studies (GWAS) using single nucleotide polymorphism (SNP) data.
[f5] Debility means frailty, poor health, or weakness, especially resulting from old age.

REFERENCES:
1. Jonsson, T., Atwal, J.K., Steinberg, S., Snaedal, J., Jonsson, P.V., Bjornsson, S., Stefansson, H., Sulem, P., Gudbjartsson, D., Maloney, J. et al. (2012) A mutation in APP protects against Alzheimer's disease and age-related cognitive decline. Nature, Advance online publication.
2. Ball, M.P., Thakuria, J.V., Zaranek, A.W., Clegg, T., Rosenbaum, A.M., Wu, X., Angrist, M., Bhak, J., Bobe, J., Callow, M.J. et al. (2012) A public resource facilitating clinical use of genomes. Proceedings of the National Academy of Sciences, USA, Advance online publication.
3. Eisenberg, D.T., Kuzawa, C.W. and Hayes, M.G. (2010) Worldwide allele frequencies of the human apolipoprotein E gene: climate, local adaptations, and evolutionary history. American journal of physical anthropology, 143, 100-111.
4. Roberts, N.J., Vogelstein, J.T., Parmigiani, G., Kinzler, K.W., Vogelstein, B. and Velculescu, V.E. (2012) The predictive capacity of personal genome sequencing. Science translational medicine, 4, 133ra158.
5. Medawar, P.B. (1952) An unsolved problem in biology. HK Lewis and Co, London.
6. Williams, G.C. (1957) Pleiotropy, Natural Selection, and the Evolution of Senescence. Evolution, 11, 398-411.
7. Vassar, R., Kovacs, D.M., Yan, R. and Wong, P.C. (2009) The beta-secretase enzyme BACE in health and Alzheimer's disease: regulation, cell biology, function, and therapeutic potential. Journal of Neuroscience, 29, 12787-12794.
8. Head, D., Bugg, J.M., Goate, A.M., Fagan, A.M., Mintun, M.A., Benzinger, T., Holtzman, D.M. and Morris, J.C. (2012) Exercise Engagement as a Moderator of the Effects of APOE Genotype on Amyloid Deposition. Archives of neurology, Epub ahead of print.