One result of the advances in genetic technology brought on by the Human Genome Project was next-generation sequencing, which can decipher huge stretches of the genome in short order, so that we can now objectively scan the entire human genome of a patient to find causative mutations underlying their particular disorder. Then it was discovered, as we mentioned before, that for most common diseases with a genetic component, only about 5 percent of the gene mutations associated with the disease are sufficient to cause it. These “fully penetrant” mutations, once inherited, guarantee the disease. (They are also called Mendelian gene mutations, after the famed pea-growing monk Gregor Mendel, the father of genetics.)
In fact, the first Alzheimer’s disease genes that Rudy and others discovered in the late 1980s and 1990s contained such mutations. However, in 95 percent of inherited diseases, variations in the DNA of numerous genes (variants) conspire with one another to ultimately determine someone’s risk for disease, adding in lifestyle habits and experience. These variants in the DNA are classified as genetic risk factors. While some increase risk, others can protect us from disease. In the majority of cases, however, the outcome depends on environmental exposure and lifestyle.
For a specific individual, discovering exactly how much contribution is being made genetically involves a huge amount of detective work, scouring multiple gene variations at once and comparing the results to the patient’s family history, life experiences, and environmental exposures. So despite the considerable success among gene hunters like Rudy and his team, for many disorders—for example, schizophrenia, obesity, bipolar disease, and breast cancer—the gene variants associated with the disease have to date accounted for less than 20 percent of the variance underlying risk.
For most complex diseases, it is now realized, there’s an interplay of nature and nurture. In this interplay, the influence of epigenetic factors assumes a major role. Epigenetics mechanisms have already been linked to many diseases, including the childhood disorders Rett syndrome, Prader-Willi syndrome, and Angelman syndrome. In some cases, gene activity is turned off directly by methylation of the bases of DNA in the gene itself. In other cases, chemical modifications (methylation and acetylation) are made to the histone proteins that bind the DNA in order to silence the gene.
But the picture has become still more complicated. Now that we can sequence whole genomes, we are finding that each of us carries up to 300 mutations that lead to the loss of function of specific genes as well as up to 100 variants that have been associated with risk for certain diseases. Moreover, some mutations and DNA variants influencing risk weren’t present in the genomes of our parents but occurred anew in the sperm or egg. These are called de novo, or novel, mutations. Novel mutations can occur in the sperm and the egg that joined to form your embryo. Such mutations occur 1.2 times every 100,000,000 bases in the two sets of 3 billion DNA bases you inherited from your parents.
That means you harbor in your genome roughly 72 de novo mutations that your parents don’t carry in their genomes. (The actual rate of de novo mutation is heavily dependent on the father’s age when the baby is conceived. Every sixteen years after the age of thirty, the number of mutations in paternal sperm doubles, which has been shown to contribute to the risk for diseases such as autism.)
In addition to single-base variants in your DNA, you carry large duplications, deletions, inversions, and rearrangements of up to millions of bases of DNA—these are known as structural variants (SVs). Like the single-base variants (technically notated as SNV, for single-nucleotide variant), structural DNA disruptions can either be inherited from your parents or occur as de novo mutations. In Alzheimer’s disease, a duplication of the APP (amyloid precursor protein) gene, the first Alzheimer’s gene to be discovered, inevitably leads to early-onset (under age sixty) dementia.
SVs and SNVs can both be found by next-generation DNA sequencing. But in another type of genetic analysis, gene expression (or gene activity) can be assessed across the entire genome. This is called transcriptome analysis. When a gene makes a protein, it first makes an RNA transcript that will be used to guide the synthesis of the protein. Transcriptome analysis can be used as part of testing for epigenetic regulation of genes, since it provides information about gene activity, not the sequence of the DNA.
The point is that powerful tools are now available to unravel the complexity of most diseases that have a genetic component. One issue is that the way a complex disease progresses is by a series of steps connected to one another. In everyday life, when you catch a cold, you first notice a mild symptom like a scratchy throat, and unless you catch the cold at this very early stage (by taking zinc tablets, for instance), you know from experience that a chain of symptoms will follow. Something similar is involved in genetics. Genetic studies using transcriptome analysis and whole-genome sequencing together carry out “pathway analysis,” which looks at many genes involved with a disease at once. With this information, the aim is to understand the pathological mechanisms by which the disease is caused and progresses. Specific biological pathways—for example, inflammation or wound healing—influence the risk for disease. Pathway analysis also elucidates other new genes of interest that might be involved in the disease, based on the biological pathways implicated. For example, in Rudy’s studies of Alzheimer’s disease, pathway analyses of the risk genes that he and others discovered have implicated a major role for the immune system and inflammation. When it comes to human disease, whether it’s cancer, diabetes, heart disease, or Alzheimer’s, to name a few, inflammation is almost always the killer that takes the patient out. If you wanted to name the epigenetic change that plays the biggest role in modulating a biological process, it would probably be inflammation.
Close to 400 million people worldwide suffer from type 2 diabetes (T2D), a number that’s expected to grow to well over 500 million in the next twenty years. In T2D patients, plasma glucose (or blood sugar) levels are elevated, often later in life as a consequence of both genetics and lifestyle choices, particularly diet. A major risk factor is obesity. One often sees clustering of diabetes in families, and while this would normally implicate gene mutations that run in the family, the family members also tend to eat together, sharing the same diet and probably similar eating habits.
Risk has become more precise but not necessarily simpler. In T2D, dozens of genes are already known to be associated with risk for adult onset. (Not surprisingly, many of these genes have also been associated with obesity and altered glucose levels.) However, most of the DNA variants in the implicated genes exert only small effects on lifelong risk for the disease. Lifestyle is probably most of the story, which you now know means that epigenetics is at work. Some of the strongest evidence for this comes from findings that a person’s early diet and nutrition in childhood determine later life risk for diabetes and heart disease. The Pima Indian population in Arizona is heavily affected by T2D and obesity. If a Pima mother was suffering from T2D while pregnant, the children turn out to be highly prone to both T2D and obesity.
The science tying epigenetics to complex disease is emerging at a frenetic pace. We now have gene chip technologies that can search through half a million sites in the genome to find where methylation may be turning off the activity of any of our 23,000 genes. These sites can be scanned for specific diseases like diabetes to ask exactly which genes are being switched. These epigenome-wide association studies, as they are called, are now being carried out around the world for all the most common disorders. In the case of T2D, some of the greatest epigenetic modifications were found around a gene called FTO, which has been linked to obesity and body mass index, which measures the ratio of fat in overall weight.
Another factor contributing to risk for diabetes is birth weight. It turns out that future risk for diabetes is highest in babies born with either low or high birth weights. Epigenetic effects on the genome of low-birth-weight babies can begin in the uterus. For high-birth-weight babies, the issue seems to be exposure to diabetes in the mother during pregnancy. All in all, the risk for T2D almost certainly involves a combination of genes, lifestyle, and epigenetics in which all these factors interplay. The same model is likely to apply for most complex diseases, from metabolic disorders to addictions and psychoses.
A field of study that has long been close to Rudy’s heart is Alzheimer’s disease. In 2015 a comprehensive analysis of the role of epigenetics in Alzheimer’s was reported in the journal Nature, and the results were striking. Researchers at the Massachusetts Institute of Technology (MIT) used mice altered with a human gene that caused them to undergo nerve cell loss, or neurodegeneration. This kind of nerve cell death is similar to what happens in the brain of a patient in the final stages of Alzheimer’s, which basically robs one of oneself.
As nerve cells started to die in the brains of the mice, the investigators looked for accompanying changes in the epigenome. As rampant neurodegeneration took over the brain, genes in two major categories were found to carry epigenetic marks. These included genes involved in neuroplasticity and the rewiring of neural networks—crucial to the brain’s ability to renew itself—along with other genes involved with the brain’s immune system. The brain’s immune system uses inflammation to protect the brain, often at the expense of nerve cells, which die in the wake of unbridled inflammation.
In the latter case, cells known as microglia, which normally support and clean up after nerve cells, sense the surrounding massacre and assume, mistakenly, that the brain is under attack by bacteria or viruses. Consequently, the hyped-up microglial cells start shooting free radicals (oxygen-based bullets) to kill the foreign invaders. In the process, they kill many more nerve cells as a sort of collateral damage in battle.
The MIT team then compared the epigenomic signature of the brains of the altered mice to the autopsied brains of Alzheimer’s patients who had succumbed to the disease. Uncanny matches were observed. (These findings were later extended to epigenetic marks in patients currently suffering from the disease.) Starting in 2008, Rudy’s group and others were increasingly finding new Alzheimer’s-associated genes functioning as part of the brain’s immune system, carrying mutations that predispose to inflammation. When the results of Rudy’s Alzheimer’s Genome Project were combined with the MIT group’s data, the message was loud and clear: Alzheimer’s is essentially an immune disease driven by the interplay of immune gene mutations and lifestyle, ultimately culminating in epigenetic alterations of those same immune genes.
An entirely new paradigm for the cause and progression of Alzheimer’s disease was being born. Rudy’s team and others are still trying to figure out how to “chill out” the brain’s immune system as a way to prevent and treat the disease. The answers will undoubtedly lie in the way immune genes are orchestrated to deal with the onslaught of neurodegeneration in the brain.
We’d like to address the intriguing trail of clues that solved one of the chief mysteries behind Alzheimer’s disease. As it turns out, sleep was one of the main clues. Disturbances in the sleep/wake cycle have been associated with numerous neurological and psychiatric diseases, including Alzheimer’s disease. Science is arriving at a pretty good idea of how sleep is linked to Alzheimer’s. We now know that the disorder is initiated by the excessive accumulation in the brain of a small protein called beta-amyloid, written variously as β-amyloid and amyloid-β (Aβ), which was not always obvious. When Rudy was a student in the mid-eighties, he and others in the field had maintained that Alzheimer’s is initiated by brain amyloid deposits. In 1986, Rudy and others discovered the gene (APP) that makes Aβ (this also turned out to be the first Alzheimer’s gene), and twenty-eight years later he and his colleagues developed the first model of Alzheimer’s pathology in a laboratory petri dish by growing brain nerve cells in an artificial brain-like environment. In that study, Rudy and colleagues Doo Yeon Kim, Se Hoon Choi, and Dora Kovacs were able for the first time to fully recapitulate the senile (amyloid) plaques and tangles inside the nerve cells that litter the brains of Alzheimer’s patients. The study earned the team a highly prestigious Smithsonian American Ingenuity Award in 2015.
The creation of “Alzheimer’s-in-a-dish,” as the New York Times dubbed it when they reported on the scientific paper in Nature announcing the achievement, settled a thirty-year debate.* That debate, in fact, had been the biggest in the Alzheimer’s field. The debate was over whether excessive amounts of amyloid surrounding the outside of affected brain cells was the actual cause of forming tangles inside the cells, leading to their death. (Tangles are an abnormal aggregate of proteins inside a brain cell that serve as a critical marker for Alzheimer’s.) The new study provided the first convincing evidence that β-amyloid can trigger all the subsequent pathology leading to nerve cell death and Alzheimer’s dementia.
Alzheimer’s is the most common cause of dementia in elderly people, and sufferers frequently experience major sleep problems. While these sleep disturbances were once dismissed as a simple consequence of the disease, we know they occur early on and may actually help cause Alzheimer’s. Considerable evidence indicates that the sleep/wake cycle is tightly linked to the production of β-amyloid in the brains of humans and mouse models of Alzheimer’s disease. As shown by Rudy’s colleague David Holtzman at Washington University in St. Louis, more amyloid is produced at higher levels in the brain when we are awake and nerve cells are more active. At night, particularly during deep sleep (slow-wave sleep), amyloid production is turned way down. Some other useful things happen in the brain during deep sleep. First, it is believed by some scientists that during deep sleep, short-term memories are consolidated into long-term memories, rather like downloading data from your thumb drive to your hard drive. Second, with regard to Alzheimer’s, not only is β-amyloid production turned down during deep sleep, but this is also the time when the brain literally cleans itself out. It produces more fluid around brain cells, which serves to flush out the bulk of metabolites and protein debris like β-amyloid. This waste-clearance pathway is referred to as the brain’s glymphatic system, resembling what the body’s lymphatic system does but employing the brain’s glial cells rather than lymph cells. So not only do you get a break from β-amyloid formation as nerve cell activity slows down during deep sleep, but you also clear it out of the brain. Meanwhile, humans or mice that are sleep deprived—a major stressor—make much more β-amyloid and show evidence of elevated nerve cell injury and even tangle pathology. Given that β-amyloid and tangles drive nerve cell death in Alzheimer’s disease, there is now an added reason to get eight hours of sleep every night and avoid the stress placed on your system by sleep deprivation. Good sleep is promising as one of the best ways to potentially lower your risk for Alzheimer’s. It’s also possible that improving the quality and duration of sleep in Alzheimer’s patients could help them. While we do not yet understand exactly how sleep cleans out the brain at the level of our genes, attending to your own sleep helps reduce the anxiety provoked by this terrible disease.
Another disease with complex patterns for risk is breast cancer. Researchers at University College London have revealed much of the epigenetic signature for breast cancer by studying healthy women who later went on to get breast cancer, with or without the presence of a mutation in the BRCA1 (pronounced “bra-ca 1”) gene. BRCA1 mutations are responsible for about 10 percent of breast cancers, leaving the other 90 percent largely a mystery. The question is, how much “missing heritability” is epigenetic? It turned out that the epigenetic alterations involved were pretty similar in both groups of women; in other words, the alterations were independent of inheriting the BRCA1 gene mutation. If the disease’s epigenetic signature is known, it can eventually be used to predict who is on the way to getting breast cancer before it strikes, a major advance given that every year 250,000 women get the disease, and 40,000 die from it.
The fact that epigenetics has such an apparently strong effect on risk means we must deeply consider lifestyle changes, beginning with diet. Among nutrients and supplements that have been validated to help reduce the risk for breast cancer are aspirin, coffee, green tea, and vitamin D.
In the case of aspirin, the best data come from a 30-year study that followed 130,000 people. Those who regularly took aspirin (at least two 325-milligram aspirin tablets per week) had a decrease in gastrointestinal cancer of 20 percent and a decrease in colorectal cancer of 25 percent. The results for these specific cancers don’t apply across the board to cancer in general, and it took 16 years of taking aspirin for the benefit to appear. If people stopped taking aspirin for 3 or 4 years, their advantage disappeared. The reason aspirin works against cancer, so far as is known, is connected to its anti-inflammatory effect (no surprise) and its apparent ability to decrease the formation of new cancer cells.
In heart disease, we also know that gene mutations and lifestyle work together to determine risk, but as in diabetes and breast cancer, so do epigenetic modifications (methylation) that silence certain genes. In one study it was found that levels of two blood fats (triglycerides and very-low-density lipoprotein [VLDL] cholesterol) were tied to methylation of a gene called carnitine palmitoyltransferase 1A (CPT1A). This gene makes an enzyme needed to break down fats. When it is turned off by epigenetic mechanisms, instead of fatty acids in the body being converted into energy, they stay around in the bloodstream, increasing the risk for heart disease. Methylation of the CPT1A gene is affected by diet, alcohol, and smoking.
Even alcohol dependence is affected by epigenetic events. Alcoholism takes a devastating toll on the victims as well as their families, contributing to 1 in 30 deaths worldwide. The best-known genes associated with alcohol dependence are alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH). Both make enzymes that help break down alcohol in the body. But variations in these genes explain only a minor degree of the inheritability of alcoholism. The “missing heritability” likely lies in epigenetic changes that are tied to the reward centers of the brain, the source of feeling good when you take a drink.
Now we know that these reward centers actually undergo changes in gene activity following the intake of alcohol. This means that different people will respond to alcohol consumption in different ways, depending on their gene activities. In heavy drinkers, an amino acid called homocysteine may go up, ultimately leading to methylation changes that silence specific genes. Such gene activities can trigger a vicious circle in which the response to pleasure and pain is altered, leading to an increased craving for alcohol to deliver less and less pleasure.
Epigenetic modifications can also be tied to psychiatric disorders like schizophrenia and bipolar disease. Finding the inherited gene mutations that lead to these illnesses has so far met with only limited success. This impasse once again leaves a potentially significant role for epigenetics in helping to fill in the missing heritability and the role of lifestyle. Increasing evidence shows that schizophrenia and bipolar disorder may not be guaranteed by, or solely dependent on, gene mutations that are passed on from parent to child.
Suspected culprits in someone’s lifestyle include diet, chemical toxins, and child rearing that affects epigenetic modifications. A patient’s lifestyle can determine epigenetic marks acquired since birth, but mouse studies would suggest that other epigenetic marks may be inherited. These marks presumably would arise as a result of the lifestyles of the parents or even grandparents. (Please note that we aren’t suggesting blame. The epigenetics of mental illness are quite tentative and incomplete. No one has yet connected A to B for any lifestyle choice that may be implicated in mental disorders.)
Epigenome-wide studies of schizophrenia and bipolar disorder have revealed epigenetic marks on some predictable genes, such as those involved with making certain neurochemicals previously associated with psychosis. But others were less predictable. For example, key genes required for immunity have turned up in both schizophrenia and bipolar disorder, suggesting that the immune system may be somehow related to a susceptibility to these disorders. Of course, here and in other epigenetic signatures associated with risk, cause and effect are an issue. How do we know whether the epigenetic marks occurred previous to onset (cause) or as a result of the disease (effect)? For now, it’s safe to say that epigenomic tests for specific diseases will become invaluable in every aspect of preventing and treating complex diseases, from prevention to ultimate cure.
In fact, we are tremendously optimistic about where genetics is leading, but we are realists, too. There remains a sharp divide between two domains, the visible and the invisible. All of us live in both domains, a fact that can’t be ignored. Peering through a microscope, a cell biologist can witness myriad changes in how a cell is functioning, yet the most crucial component, the experience that guides these changes, cannot be observed. The nonphysical is playing its part during every second of a person’s life, and we believe it’s the prime reason why genetics must look beyond materialism and random chance.
The data will have to support such a radical change in perspective, but far more important is to formulate the ideas that the data must fit—that’s our objective in this book, and we’ve taken some giant steps in that direction. You now know more about the dynamic nature of your genome than geneticists knew even twenty or thirty years ago. What’s most crucial, however, is applying the knowledge to optimize your genetic activity. Before we can do that, another big chunk of genetic information needs to be presented, and it comes from a very surprising source that no one ever anticipated.
* The Alzheimer’s-in-a-dish study was made possible by a very forward-looking foundation, the Cure Alzheimer’s Fund.