The Four Known Alzheimer’s Genes

Over the past several decades, it has become increasingly clear that inheritance plays a major role in Alzheimer’s disease. The roughly 25,000 genes in the human genome are comprised of deoxyribonucleic acid (DNA) packaged into 24 different chromosomes, 1-22, X and Y. A gene’s job is to either make proteins or control the activity of other genes. Over many generations, the DNA of a gene can mutate to create a “variant”. A very rare DNA variant is called a “mutation”, while a variant that is common in the population is called a “polymorphism”. DNA variants allow for all of us to be a little different from each other. There are about 3 million variants that differ between any two individuals. Variants in certain genes can directly cause a disease like Alzheimer’s, can increase susceptibility to disease, or can even confer protection against disease. One’s risk for most age-related diseases such as cancer, diabetes, heart disease, stroke, and Alzheimer’s is strongly influenced by our genes. For all of these age-related diseases, we know of mutations that guarantee onset of these diseases with no need for input from any other genes or environmental factors. And, we know of polymorphisms that can increase (or decrease) one’s susceptibility to the disease, but without guaranteeing onset of the disease. In this latter case, other genes and environmental factors usually conspire together to determine when and whether one will get disease. Any gene which can contain a variant(s) that significantly influence one’s susceptibility to Alzheimer’s, whether it be to guarantee the disease or serve to increase (or decrease) risk, is called an “Alzheimer’s gene”. It is important to remember that all genes are “good”; it is only the variants in the DNA of these genes that can influence one’s lifetime risk for a disorder such as Alzheimer’s disease.

In the 1980’s and 90’s, my laboratory co-discovered the three known genes that can carry mutations causing early-onset (<60 yrs) familial Alzheimer’s disease. These three genes, known as APP, PSEN1 and PSEN2, can harbor any of over 200 different gene mutations that guarantee onset of Alzheimer’s at a relatively early age with no need for additional input from other genes or environmental factors. These mutations are rare, accounting for only 1-2% of Alzheimer’s cases. Inheritance of one of these mutations from just one parent virtually guarantees onset of Alzheimer’s, usually by 60 years old. If a parent carries such a mutation, each child has a 50% chance of inheriting the same mutation and getting early-onset Alzheimer’s disease with virtual certainty before 60 years old. Genetic testing is available for the early-onset Alzheimer’s gene mutations, but is usually reserved for those who have a family history of early-onset Alzheimer’s disease.

The fourth known Alzheimer’s gene is APOE. In the early 1990’s, investigators at Duke University found that a common gene variant (polymorphism) of APOE, called epsilon 4, can increase risk for late-onset (>60 yrs) Alzheimer’s disease. This variant is present in about 20% of the general population but this increases to >50% in Alzheimer’s patients. Unlike the early-onset AD gene mutations, this variant does not guarantee Alzheimer’s, but only serves to increase risk. Inheriting one copy of the variant (from one parent) increases risk by 4-fold (versus the general population) and two copies (from both parents), >10-fold. Importantly, a person can inherit the APOE epsilon 4 gene variant from one or both parents and never get Alzheimer’s in the span of a normal lifetime.

With regard to genetic testing for the common late-onset form of Alzheimer’s, we are not yet able to do so reliably. This is because the APOE epsilon 4 gene variant is not sufficient on it’s own to predict one’s risk for Alzheimer’s reliably. Other genes and environmental factors need to combine with the APOE epsilon 4 gene variant to cause Alzheimer’s. Some gene variants can exacerbate while others mitigate the risk for Alzheimer’s conferred by the APOE epsilon 4 gene variant. And, we do not yet know the full set of gene variants that can increase or decrease risk for Alzheimer’s when inherited together with the APOE epsilon 4 gene variant. Thus, APOE gene testing is not recommended as a sole means for predicting Alzheimer’s risk. The other late-onset Alzheimer’s genes must first be identified in order to reliably test for risk for late-onset Alzheimer’s disease.

So how many other Alzheimer’s genes are there? We know that the four known Alzheimer’s genes, APP, PSEN1, PSEN2, and APOE account for roughly 30% of the inheritance of Alzheimer’s. Thus, 70% of the genetics of Alzheimer’s remains undefined. We as well as others have been engaged in comprehensive projects to find the other Alzheimer’s genes. Once we have all of the Alzheimer’s genes in hand, we will be able to more reliably predict one’s lifetime risk for the common late-onset form of Alzheimer’s disease. However, one might ask, “Why bother to test if there is nothing we can currently do to prevent, stop, or reverse it?” This is certainly a fair question since we still do not have drugs that stop the disease process in Alzheimer’s. We only have drugs like Aricept and Namenda that modestly and temporarily alleviate the symptoms of cognitive decline, but without affecting the progress of the disease.

We need to do better more effective therapies for Alzheimer’s, but how do we get there? First we need to identify all of the genes and variants involved in influencing risk for Alzheimer’s disease. Studies of the known Alzheimer’s disease genes (the four previously mentioned) have provided the vast majority of information being used to guide novel drug discovery aimed at preventing, stopping and maybe even reversing Alzheimer’s disease. Every new Alzheimer’s gene defect we find provides new clues regarding the cause of the disease what we need to do to stop the disease. Thus far, all four genes have pointed to a small protein called “Abeta” as the cause. Abeta is normally made in the brain, but is found in excessive amounts in the brains of Alzheimer’s patients, e.g. in senile plaques that litter the Alzheimer’s brain around nerve cells. Small clumps of Abeta can gum up the connections between nerve cells known as synapses. Billions of nerve cells in the brain form trillions of synapses making up our neural network. The neural network, in all its complexity, is needed for all brain function, including memory and learning. Excessive Abeta disrupts synaptic communication between nerve cells leading to loss of memory and learning and eventually dementia. Dementia is defined as global and catastrophic cognitive failure; Alzheimer’s disease is the most common form of dementia in the elderly.

Beyond the Original Four Alzheimer’s Genes

Most drug discovery for Alzheimer’s today is based on studies of the four original Alzheimer’s genes. But, we know that there are many more Alzheimer’s genes yet to be identified. Since 2005, the Cure Alzheimer’s Fund has supported a project called the Alzheimer’s Genome Project (AGP), carried out in my laboratory at Massachusetts General Hospital. The goal of this project is to study 5000 families with multiple members who are affected with the common late-onset form of Alzheimer’s disease in an effort to identify all of the other Alzheimer’s genes. In addition to the Alzheimer’s Genome Project, the International Genomics of Alzheimer’s Project (IGAP a consortium of dozens of research institutions in Europe and the U.S., in which we are members), uses tens of thousands of individual Alzheimer’s cases from the general population in the U.S. and Europe, to find common DNA variants that influence risk for Alzheimer’s. The family-based method of our Alzheimer’s Genome Project and the population-based method of IGAP have identified some of the same Alzheimer’s genes, but also find different ones with different effects on risk.

In the family-based studies of the Alzheimer’s Genome Project, we are able to find not only common DNA variants that influence one’s risk for Alzheimer’s, but also rare mutations that profoundly affect risk or for the disease or directly cause it. The Alzheimer’s Genome Project places a high priority on finding these rare but very potent gene mutations because historically, amongst the four known Alzheimer’s genes, it has been the rare early-onset familial Alzheimer’s gene mutations that have been most effectively guiding drug discovery efforts. This is mainly because hard-hitting mutations have clear-cut adverse effects on biological systems, which can be elegantly recapitulated in animal models. This then allows for more effective drug discovery and development.

Along these lines, one of the first new Alzheimer’s genes to be identified in the Alzheimer’s Genome Project was ADAM10. This gene was specifically chosen for testing as a potential Alzheimer’s gene because like the four original Alzheimer’s genes, it affects the production of Abeta in the brain. We identified two rare mutations in this gene that strongly predispose carriers to Alzheimer’s disease at around 70 years old. These two mutations were found in only 7 (of 1000 AD families tested). Thus, they are very rare. We have recently demonstrated these two mutations dramatically impair the activity of ADAM10. ADAM10 normally blocks the production of Abeta. Accordingly, we have found that these two rare mutations greatly enhance Alzheimer’s amyloid pathology in animal-based models of the disease. With the validation of these mutations to be “pathogenic”, or disease causing, the Alzheimer’s Genome Project considers ADAM10 to be the fifth Alzheimer’s gene. We published the original findings showing ADAM10 to carry rare mutations causing Alzheimer’s disease in 2008 in the prestigious scientific journal, Human Molecular Genetics. The publication of the validation data from transgenic mouse models is planned for the coming year in 2011.

Also beginning in 2005, with Cure Alzheimer’s Fund support, as part of the Alzheimer’s Genome Project, we carried out the first family-based “genome-wide association study” for new Alzheimer’s genes. This entailed a screen of the entire human genome in patients and their relatives in thousands of Alzheimer’s families. The first phase of this study was completed in 2008 and led to the identification of over 100 new Alzheimer’s candidate genes. We reported the top four Alzheimer’s candidate genes from this study in 2008; it was named by TIME/CNN to be one of the Ten Top Medical Research Breakthroughs of 2008.

The Cure Alzheimer’s Fund Alzheimer’s Genome Project was the first large-scale study of the human genome performed in the world on the world’s largest collection of families affected by Alzheimer’s disease. It was also the first genome-wide study for Alzheimer’s in the world to discover novel Alzheimer’s gene candidates with statistically significant results and confirmation in thousands of subject from families with a high incidence of Alzheimer’s disease. The four new Alzheimer’s genes reported by the Alzheimer’s Genome Project in 2008 in the American Journal of Human Genetics included: ATXN1, CD33, GWA14Q34, and DLGAP1. ATXN1 is known to carry mutations that cause another neurodegenerative disease called spinal cerebellar ataxia, a movement disorder. We found that when this gene is inactive, Abeta levels increase dramatically leading to cognitive decline in mouse models. Another gene, CD33, is perhaps the most interesting since it controls the brain’s innate immune system and inflammation in the brain. In a related study funded by the Cure Alzheimer’s Fund, Abeta was found to play a role in the brain’s innate immunity system. CD33 regulates the brain’s immune system and concurrently, levels of Abeta. We are now developing CD33 as a drug target for Alzheimer’s based on the genetic findings of the Alzheimer’s Genome Project. It should be emphasized that without the Cure Alzheimer’s Fund Alzheimer’s Genome Project we would probably have never guessed that genes like ATXN1 and CD33 might be involved with Alzheimer’s.

As part of the current activities of the Alzheimer’s Genome Project, we are now testing these as well as over one hundred other new Alzheimer’s candidate genes, coming out of our genome screen, to identify all of the DNA variants and mutations that influence risk for Alzheimer’s in the five thousand Alzheimer’s families under study in the Alzheimer’s Genome Project. We are specifically searching for DNA mutations and variants in these genes that very strongly affect risk for onset of Alzheimer’s. As new defects are found in these genes, we will not only increase our ability to reliably predict risk for Alzheimer’s, but more importantly, garner new clues regarding the causes of Alzheimer’s, and in doing so, gather new ideas and biological targets for novel drug discovery aimed at preventing, stopping and reversing Alzheimer’s disease.

In the parallel screen for new Alzheimer’s genes conducted by the IGAP, the DNA from tens of thousands of individual Alzheimer’s patients was compared to the DNA of elderly subjects without Alzheimer’s to find common variants that influence risk for Alzheimer’s. In 2009, this led to the identification of four new Alzheimer’s gene candidates called PICALM, CLU, CR1, and BIN1. More recently in April 2011, IGAP found four more Alzheimer’s genes called CD2AP, MS4A, EPHA1, and ABCA7. In addition, they found Alzheimer’s risk to be influenced by the gene CD33, which was first reported by our Alzheimer’s Genome Project in 2008.

It should be noted for the sake of clarity that the IGAP had stated in their reports and press releases that they had increased the number of known late-onset genes from “five to ten”. However, these numbers only pertained to studies of individual Alzheimer’s patients in the general population screened in IGAP, and not the family-based Alzheimer’s genes reported by the Alzheimer’s Genome Project. The IGAP considered the original five late-onset Alzheimer’s genes to be APOE (discovered as mentioned earlier at Duke U. in the early 1990’s), PICALM, CLU, CR1, and BIN1. They then considered the next five to be CD2AP, MS4A, EPHA1, ABCA7, and CD33. However, as mentioned above, CD33 had been already identified earlier in our Alzheimer’s Genome Project in 2008, which was reported in the major scientific journal, The American Journal of Human Genetics. In the IGAP announcement, they also overlooked the other three late onset genes, which had been discovered three years earlier in our Alzheimer’s Genome Project. So, in fact, with the 8 new genes reported by IGAP and the 5 new genes reported by the AGP, there have been 13 new late onset Alzheimer’s genes discovered in the last 5 years, which, when added to the discovery of APOE yields 14 total late onset genes now reported in the scientific literature. To that total would be added the 3 early onset genes co-discovered by Dr. Tanzi and colleagues to reach a total of Alzheimer’s genes discovered of 17. In addition, Dr. Tanzi and AGP have also identified over 100 unpublished Alzheimer’s candidate genes that are currently being confirmed and validated for publication over the coming year.

With regard to effects on risk, all of the new Alzheimer’s gene candidates reported by the IGAP carry common DNA variants that confer only tiny effects on risk. Specifically, according to IGAP, the new genes contain common DNA variants that are present in a large proportion (30-70%) of the general population, but only increase or decrease risk for a given individual by a mere 10-20%. In contrast, the epsilon 4 variant in APOE, which is present in 20% of the population, increases risk by 400 – 1200 %! And the ADAM gene just discovered by the AGP increases risk for the individuals who have it by about 500%

With regard to the CD33 gene, which was identified as an Alzheimer’s gene in both our Alzheimer’s Genome Project in 2008 and the IGAP in 2011, each project actually discovered different Alzheimer’s-associated DNA variants in this gene. In our family-based Alzheimer’s gene study, we originally reported a relatively uncommon variant in CD33 that increases risk for Alzheimer’s in a subset (<100) of the 5000 Alzheimer’s families we studied. In contrast, the IGAP discovered a very common variant in CD33, present in about 50% of the population that conferred only marginal protection against Alzheimer’s (decreasing risk by only 11%). The fact that we now know of two different Alzheimer’s-associated DNA variants in the CD33 gene from multiple Alzheimer’s samples increases the odds that CD33, is a bona fide Alzheimer’s gene.

As with all of the new genes found in the genome-wide association screens of the Alzheimer’s Genome Project and IGAP, the next critical step is to identify all of the DNA variants and mutations in these genes that increase or decrease risk for late-onset Alzheimer’s disease. The Cure Alzheimer’s Fund continues to support these efforts. We are currently screening over a 100 new Alzheimer’s candidate genes found in the Alzheimer’s Genome Project along with those found in the IGAP, to identify the all of DNA variants and mutations in these genes that influence risk for Alzheimer’s disease. Elucidating the full deck of Alzheimer’s-associated gene variants and mutations and understanding the interrelationships among them is is necessary to fully understand all of the biological processes that are affected in Alzheimer’s disease. This will give us the best odds of reliably predicting the disease early in life (with appropriate counseling and legal protection). But, most importantly, the full set of Alzheimer’s genes and the knowledge of how they biologically influence risk for disease will continue to provide the most critical information needed to guide the development of new and effective therapies aimed at preventing, stopping or reversing Alzheimer’s disease.

Finally, it should be noted that whether a DNA mutation in an Alzheimer’s gene is rare and restricted to a small subset of families or more broadly observed in the general population, most believe that new drugs or therapies for Alzheimer’s based on what is learned from that mutation will be useful in preventing and treating all cases of Alzheimer’s. As noted above, in the Alzheimer’s Genome Project, we place a high priority on family-based gene studies of Alzheimer’s since there we have the highest odds of finding DNA mutations with very strong effects on risk for Alzheimer’s, akin to those of the early-onset familial Alzheimer’s disease gene mutations discussed above. These mutations are most useful for driving successful drug discovery since their biological effects on the disease process are of much greater impact and more clear-cut in terms of mechanism by which they cause disease. They also lend themselves to more useful animal models for drug testing. Ultimately, the full list of Alzheimer’s genes emerging from the family-based genetic studies of the Alzheimer’s Genome Project and the population-based studies of IGAP getting us closer and closer to someday being able to eradicate Alzheimer’s disease using a strategy of early prediction and early intervention.

Alzheimer’s Genes Identified To Date (Total of 17):

Early-onset familial Alzheimer’s disease genes (onset < age 60):

APP (1987)*
PSEN1 (1995)*
PSEN2 (1995)*

Late onset genes (onset > age 60):

APOE (1993)
ADAM 10 (2008)**
ATXN1 (2008)**
CD33 (2008)**
GWA14Q34 (2008)**
DLGAP1 (2008)**
PICALM (2009)***
CLU (2009)***
CR1 (2009)***
BIN1 (2010)***
CD2AP (2011)***
MS4A (2011)***
EPHA1 (2011)***
ABCA7 (2011)***

* Co-discovered in Tanzi laboratory

** Discovered by the Cure Alzheimer’ Fund Alzheimer’s Genome Project in Alzheimer’s families-these genes are expected to contain DNA variants that significantly increase risk for Alzheimer’s.

*** Discovered by the IGAP-these genes are expected to contain DNA variants that are common in the general population but which have only tiny effects on risk for Alzheimer’s.

N.B. The Alzheimer’s Genome Project has also discovered over 100 additional Alzheimer’s candidate genes that are in the process of being confirmed and validated. In addition, the Cure Alzheimer’ Fund supports a website called AlzGene Http://alzgene.org, in which we are tracking all of the Alzheimer’s candidate genes reported in the scientific literature, including their ongoing testing for confirmation as bona fide Alzheimer’s disease genes.

With regard to functional effects, the above list of 17 Alzheimer’s genes can be divided into four major categories based on their known or predicted biological effects on Alzheimer’s risk:

1. The production and clearance of Abeta, the major protein in beta-amyloid deposits in the brains of Alzheimer’s patients (APP, PSEN1, PSEN2, APOE, ADAM10, ATXN1, CD33, CLU)
2. Cholesterol metabolism (APOE, CLU, ABCA7)
3. The innate immune system and inflammation (APOE, CD33, CLU, CR1)
4. Cell signaling and protein trafficking (PICALM, BIN1, EPHA1, CD2AP)