Cover Story | Heart Trek: Next Gen: Genes, Epigenetics—and the Final Frontier?
Although similar to genetic features of DNA in terms of their heritability, epigenetic mechanisms differ in their potential reversibility by environmental/nutritional factors, which could make them crucial in complex and multifactorial diseases like CVD. From a molecular standpoint, epigenetics refers to chromatin-based mechanisms that help orchestrate gene regulation and expression, but do not involve changes to DNA sequencing. Think of epigenetics as the crossroads between environmental factors and inheritable genetic variation. We know environment greatly impacts CV risk—and epigenetics is ground zero. With today's advanced understanding of many important genes, epigenetics should help put the genetic puzzle pieces together to assist in better treatment and prevention.
However, we are just now opening the box to see what those pieces look like. "We're really still at the early stages in CVD in terms of understanding epigenetic modifications that contribute either to disease initiation or progression," said Donna Arnett, PhD, of the University of Alabama at Birmingham, in an interview with CardioSource WorldNews.
Dr. Arnett was the first epidemiologist to be president of the American Heart Association and—as revealed during the last month of her term in May 2013—the first stroke survivor to lead the organization in its nearly 90-year history. (Her stroke occurred when she was 27.) In parallel with her work in epigenetics, Dr. Arnett remains actively involved in deciphering the cardiovascular-relevant parts of the human genome.
Despite its initial promise, and intense ongoing research, the clinical utility of genetic markers for prediction and prevention of heart disease has proved to be limited—so far. Markers are important points, but the real story is what influences those markers, which required turning attention to epigenetics as the new frontier for risk stratification, prevention, and treatment.
"The Human Genome Project has sequenced the 3 billion base pairs of our DNA," said Daniel Levy, MD, director of the Framingham Heart Study and a professor at Boston University Medical School, in an interview with CSWN. "While we've learned a lot through sequencing about how changes in DNA base pairs can affect traits and the susceptibility to disease, despite all of our understanding of genetic variation, a substantial majority of variation in human traits remains unexplained and appears not to be due to differences in the sequence of our DNA but to other factors, including things like environmental factors and epigenetic factors."
At a 2008 meeting at the Cold Spring Harbor Laboratory research institute, an operational definition of epigenetics was developed that reads: "An epigenetic trait is a stably heritable phenotype resulting from changes in a chromosome without alterations in the DNA sequence."1 The requirement for a modification to be heritable, however, has been contested as not all epigenetic modifications are transgenerational.
Epigenetic modifications fall into three distinct but interconnected processes: DNA methylation; translational histone modifications; and RNA-based mechanisms, including noncoding RNAs. Given that, Dr. Arnett explains, ‘‘epigenetic'' refers to variations in these three categories—which are all involved in CVD pathogenesis—while ‘‘genetic'' refers to polymorphisms in the DNA sequence.
Epigenetic modifications are critical both for normal human development and in offering a means for understanding cellular perturbations in the pathogenesis of disease, specifically some types of cancer, as well as in normal aging.2 The epigenome reacts to signals from the outside world such as diet or stress, allowing cells to respond dynamically to the outside world.
DNA methylation—the covalent chemical attachment of a methyl group to a nucleotide—is the most common and the most studied of the epigenetic marks. The methyl group is positioned in the major groove of the DNA, where it can easily be detected by proteins interacting with the DNA. Therefore, methylation adds extra information to the DNA that is not encoded in the sequence.
Most DNA methylation occurs within the cytosine-phosphate-guanine (CpG) nucleotides, of which about 30 million exist in a methylated or unmethylated state. Adding a methyl group at a CpG site, many of which are clustered in CpG islands, converts cytosine—one of the four main nucleobases found in DNA and RNA—to 5-methylcytosine.
"DNA methylation is a natural process by which certain positions in our DNA—and there are hundreds of thousands of those positions—can become selectively methylated and that in turn changes the machinery that generates RNA and the ability of that RNA to code for proteins, the building blocks of our whole protein-based biology," noted Dr. Levy.
He explained further that methylation in key positions in the DNA can result in changes in gene expression, resulting in more or less protein than would ordinarily be the case.
Generally, a genomic region with methylated DNA becomes inaccessible to transcriptional machinery and gene expression is suppressed. However, methylation also can be seen in a more positive light, said Dr. Arnett. In the case of cancer, for example, up-regulation of tumor suppression genes or DNA repair genes leading to greater protein expression would be a plus. Indeed epigenetic silencing of DNA repair genes has been implicated in several sporadic cancers.3
"We need the ability to up-regulate or down-regulate proteins under certain constraints of our environment. For example, in tumor progression, wouldn't it be great if things were methylated and then you shut down expression of certain proteins that stopped tumor growth?" said Dr. Arnett.
"The hope is that insights into DNA methylation will provide insights into the nexus between environmental exposure and disease, in particular with response to things like dietary changes, cigarette smoking, and alcohol intake, all of which have been shown to affect DNA methylation," added Dr. Levy.
Cancer First, Now CVD
In the early days of epigenetic research, those looking at DNA methylation focused specifically on cancer, where researchers have achieved translational success. More recently, the field of cardiovascular epigenetics has taken off; not yet to warp speed, but still advancing from global methylation studies to candidate-gene methylation studies to epigenome-wide studies. The latter utilize "easy, off-the-shelf methods driven by chip technologies" that can measure DNA methylation at approximately 450,000 CpG sites across the genome, said Dr. Arnett.
According to a review article by Dr. Arnett's group soon to be published in Translational Research, 353 of the 8,026 (4.4%) manuscripts published in 2013 relating to epigenetics concentrated on CVD.4 However, only 125 reported findings in humans; of those, 66 were reviews rather than original research. However, four were clinical trials or population-level studies, so work is moving into human studies. "We're just now reaching the cohort studies where we have big enough samples to really understand the major genes that are affected by epigenetic modifications in populations," said Dr. Arnett.
Epigenetic research and its interpretation are complicated by several factors. Unlike genetic variation, epigenetic tags are cell-specific, usually stable but also reversible, and susceptible to inherited and environmental influences.
Once it's determined which genes and epigenetic modifications are of interest, there are multiple remaining issues inherent in epigenetic investigation: the problem of reverse causality ("Does the tag cause the disease, or vice versa?"), the changing nature of epigenetic modifications, and the fact that these changes are thoroughly confounded by multiple variables including age, genotype, and lifestyle factors.
So, while we know from animal studies that cigarette smoking causes epigenetic changes almost from the first puff, once the habit is kicked, Dr. Arnett said, "we don't know how long those epigenetic modifications take to revert back, if they ever do, to their premethylated state." Why? Well, for the simple reason that it might present an ethical challenge to take a nonsmoker, have them smoke (and maybe become addicted) in order to see what happens, then see if or when their epigenome reverts to a healthier state.
At this point, most of the evidence showing that methylation patterns can revert back to a premethylated state comes from animal studies, but Dr. Arnett's group is hoping to get funding for a follow-up of a 10-year-old study that looked at epigenetic changes in response to a high-fat meal. "I'm hoping to bring them back and repeat the whole study and see if their methylation profile has reverted back to a healthier state in the 10 years since we saw them last."
Also unknown: whether epigenetic changes reflected in serum are consistent with changes in particular organs or across different cell types. That's a basic question requiring an answer, Dr. Arnett stressed, because if clinicians find useful information in the blood, they won't need to try and get tissue biopsies from organs like the heart, kidney, or brain, which clearly is not something patients will give easily.
Where No One Has Gone Before
Notwithstanding continued challenges, epigenetics is a burgeoning field of research, according to Dr. Levy, with new investigations appearing monthly that improve our general and specific understanding of this extraordinary field.
In June 2014, Dr. Arnett's group published an epigenome-wide association study (EWAS) that looked at whether there might be differential methylation of CpGs related to lipid phenotypes.5 The team identified a "robust association" between the methylation status of four markers of the gene CPT1A, the expression of CPT1A, and levels of plasma triglycerides and very low-density lipoprotein cholesterol.
While genetic loci validated as associating with lipoprotein measures do not account for a large proportion of the interindividual variation in lipoprotein measures, methylation at the CPT1A locus has been linked to lipoprotein subtraction profiles6 and CPT1A encodes the carnitine palmitoyltransferase enzyme that controls fatty acid flux in the liver, offering strong biological plausibility for the observed associations.
"This mitochondrial gene seems to be a master regulator of many other genes and to determine a lot of different metabolic effects," said Dr. Arnett. "We've also shown it to be hypermethylated in obesity, but we really don't know yet what caused the methylation that leads to all of these metabolic changes."
This started with the Genetics of Lipid Lowering Drugs and Diet Network (GOLDN) in a study of 991 participants and then replicated by Dr. Levy and colleagues in 1,261 participants in the Framingham Heart Study Offspring cohort.
"We were a replication cohort, but we went a step beyond that and were able to confirm that the particular site of DNA methylation that was related to lipid levels also affected the expression of the CPT1A gene that's involved in lipid regulation," said Dr. Levy. "So, in addition to serving as a replication cohort for that wonderful study, we provided some mechanistic insight to help understand how methylation affects lipids, in this case through the altered expression of a key gene."
Other groups have identified loci associated with high-density lipoprotein cholesterol in the setting of familial hypercholesterolemia7, and distinct patterns of DNA methylation and histone-3 lysine-36 trimethylation in left ventricular tissue of patients (and controls) with cardiomyopathy.8,9
Perhaps most exciting from a public health perspective: emerging evidence linking both DNA methylation and post-translational histone methylation to genes related to arterial hypertension.10 Moreover, DNA hydroxymethylation appears to be modifiable by changes in salt intake in a rat model, although this has yet to be replicated in humans.
Besides looking at specific CVD phenotypes, epigenetic studies have shown differential methylation in various genetic loci associated with CVD risk factors, including age, obesity, air pollution, sunlight exposure, measures of fasting insulin and insulin resistance, and smoking.
Examples abound: in July 2014, a Harvard-based group demonstrated that temperature and relative humidity were associated with DNA methylation of genes related to coagulation, inflammation, cortisol, and metabolic pathways.11 A few months prior to that, another group, this one from the United Kingdom, showed how increased body mass index is associated with increased methylation at the HIF3A locus in blood cells and in adipose tissue.12
These studies "lend support to the view of epigenetic changes as the process by which environmental factors influence inheritable genetic variation," according to Dr. Arnett and colleagues' review.3 And yet, these studies don't explain the temporal nature of the changes or show causality. And no EWAS study to date has looked at hard CVD endpoints like myocardial infarction, stroke, or cardiovascular death.
"On balance, the clinical potential of cardiovascular epigenetics has a long road to fulfillment," said Dr. Arnett and her colleagues. For her part, Dr. Arnett speculated that it will take about 5 years to have a good catalog of the genes that are most important from an epigenetic standpoint and another 5 to "understand the windows of epigenetic modification both in terms of hypo- and hyper-methylation and the natural history of those events, at which point we'll have therapeutic targets to test." And—in the words of Captain Jean Luc Picard—engage.
- Berger SL, Kouzarides T, Shiekhattar R, Shilatifard A. Genes Dev. 2009;23:781-3.
- Kumar V, Abbas AK, Fausto N, Aster J. Robbins & Cotran Pathologic Basis of Disease. 8th edition, Saunders 2009.
- Lahtz C, Pfeifer GP. J Mol Cell Biol. 2011;3:51-8.
- Aslibekyan S, Claas SA, Arnett DK. Transl Res. 2014 April 8. [Epub ahead of print]
- Irvin MR, Zhi D, Joehanes R, et al. Circulation. 2014 June 11 [Epub ahead of print]
- Frazier-Wood AC, Aslibekyan S, Absher DM, et al. J Lipid Res. 2014;55:1324-30.
- Guay SP, Voisin G, Brisson D, et al. Epigenomics. 2012;4:623-39.
- Haas J, Frese KS, Park YJ, et al. EMBO Mol Med. 2013;5:413-29.
- Movassagh M, Choy MK, Knowles DA, et al. Circulation. 2011;124:2411-22.
- Friso S, Carvajal CA, Fardella CE, Olivieri O. Transl Res. 2014 July 14. [Epub ahead of print]
- Bind M-A, Zanobetti A, Gasparrini A, et al. Epidemiology. 2014;25:561-9.
- Dick KJ, Nelson CP, Tsaprouni L, et al. Lancet. 2014;383:1990-8.
< Back to Listings