What Are Epigenetic Marks and How Do They Work?

Epigenetic marks are chemical tags attached to your DNA or to the proteins that package it, and they control which genes are switched on or off in any given cell without changing the underlying genetic sequence. The two best-known types are methyl groups added directly to DNA and a variety of chemical modifications stuck onto histone proteins (the spools around which DNA is wound). Together, these marks explain how a liver cell and a brain cell can carry identical DNA yet behave in completely different ways. The system is far more dynamic than early researchers expected, responding to diet, stress, aging, and disease in ways that are only now being mapped in detail.

DNA Methylation

The most studied epigenetic mark is DNA methylation. Enzymes called DNA methyltransferases attach a small methyl group (a carbon atom bonded to three hydrogens) to the cytosine base in DNA, almost always at spots where cytosine sits next to guanine, known as CpG sites. When methyl groups pile up in the regulatory region near the start of a gene, that gene tends to go quiet. The cell’s transcription machinery has a harder time accessing methylated DNA, so the gene’s instructions are effectively muted.

Not all CpG sites matter equally. Researchers who built a library of promoters with different densities of CpG sites found a tight correlation between CpG content and the speed at which methylation silenced a gene. They also identified a single CpG site between two key landmarks in the promoter that accounted for a large share of the difference in silencing rates, showing that certain positions carry disproportionate weight.¹

Methylation is not a one-way street. A family of enzymes called TET proteins can oxidize the methyl group, effectively beginning the process of removing it.² The TET enzymes produce intermediate forms of modified cytosine, and then a separate repair enzyme recognizes those intermediates and swaps them out for a clean, unmethylated cytosine.³ The result is active demethylation, which means cells can reactivate silenced genes when circumstances change. Three versions of TET enzymes (TET1, TET2, and TET3) safeguard low-methylation states at important regulatory regions across the genome.⁴

Histone Modifications

DNA does not float freely in the nucleus. It wraps around clusters of histone proteins, forming bead-on-a-string structures called nucleosomes. The tails of these histone proteins stick out and can be tagged with dozens of different chemical groups. Acetyl groups, methyl groups, phosphate groups, and ubiquitin tags are among the most common. Each modification alters how tightly or loosely the DNA is packaged, and by extension, how accessible a gene is to the cell’s reading machinery.

Acetylation is one of the clearest examples. Enzymes called histone acetyltransferases (HATs) add acetyl groups to specific amino acids on histone tails, which loosens the chromatin structure and tends to activate nearby genes. The opposing team, histone deacetylases (HDACs), strip those acetyl groups off, tightening the structure and suppressing gene activity.⁵ This push and pull between HATs and HDACs acts as a real-time volume dial for gene expression, and cells use it constantly.

The picture is richer than just “on or off.” Different combinations of histone marks create distinct states across the genome. Active enhancer regions carry one set of marks (acetylation of a lysine at position 27 on histone H3, along with a single methyl group at position 4). Active gene promoters carry a different signature, with high levels of trimethylation at H3 position 4 and broad acetylation across histones H3 and H4. Repressed genes, meanwhile, are marked by trimethylation at H3 position 27 or trimethylation at H3 position 9.⁶ The specific combination matters more than any single mark, and researchers sometimes describe this layered system as a “histone code.”

Chromatin Remodeling and Histone Variants

Beyond chemical tagging, cells have another trick for managing gene access: physically moving or replacing nucleosomes. ATP-dependent remodeling complexes use energy to slide nucleosomes along the DNA strand, swap out histones, or evict nucleosomes entirely to expose the underlying sequence.⁷ This is less a mark in the traditional sense and more of a structural rearrangement, but it works hand-in-hand with histone modifications and DNA methylation to determine which stretches of the genome are open for business.

Cells can also replace standard histones with specialized variants at particular locations. Swapping in a variant histone changes the physical properties of the nucleosome, which can make DNA more or less accessible. Structural studies using high-resolution imaging have begun to reveal exactly how these variant-specific features reshape chromatin.⁸ Together, remodeling and variant swapping give cells a mechanical toolkit that complements the chemical marks on DNA and histones.

Non-Coding RNAs as Epigenetic Guides

Some epigenetic marks are directed into position by RNA molecules that do not code for proteins. The most dramatic example in mammals is X-chromosome inactivation. Female cells carry two X chromosomes but only need one active copy. A long non-coding RNA called XIST coats one of the two X chromosomes, recruiting protein complexes that lay down repressive histone marks and DNA methylation until the entire chromosome is silenced.⁹ This ensures balanced gene dosage between sexes. The silenced X chromosome remains coated in XIST and densely methylated throughout a woman’s life, which is why it shows up under a microscope as a compact blob called a Barr body.

Plants use a related strategy called RNA-directed DNA methylation, in which small interfering RNAs guide methyltransferases to specific genomic locations, particularly jumping genes (transposable elements) that could wreak havoc if left unchecked. This pathway provides stable, heritable silencing of those mobile elements across generations.¹⁰ Plants also use this RNA-guided methylation to respond to pathogens and abiotic stress, making it central to how they adapt to their environment.¹¹

How Epigenetic Marks Guide Development

Every human begins as a single fertilized egg that carries a particular set of epigenetic marks inherited from the sperm and egg. Almost immediately, the embryo strips most of those marks away in a massive wave of demethylation, essentially wiping the epigenetic slate nearly clean.¹² A fresh set of marks is then laid down as cells begin to specialize. This two-wave cycle actually happens twice during the creation of each new generation: once in the cells that will become sperm or eggs (primordial germ cells), and once after fertilization in the early embryo.¹³

As cells commit to becoming a particular tissue type, the genes that maintain a cell’s ability to become anything (pluripotency genes) are gradually buried under repressive marks, including DNA methylation that locks them into long-term silence.¹⁴ Meanwhile, tissue-specific genes pick up activating marks. A developing heart cell, for instance, opens up chromatin around heart-muscle genes while silencing brain-specific and liver-specific ones. The result is that every cell type in the body carries a distinctive epigenetic fingerprint on top of identical DNA.

Even within a single lineage, epigenetic marks continue to fine-tune behavior. Blood-forming stem cells in the bone marrow, for example, appear to use epigenetic programming to develop biases toward producing certain blood cell types, and evidence suggests this diversification begins after the stem cells settle into the bone marrow rather than earlier in fetal development.¹⁵

How Diet and Stress Reshape the Marks

Epigenetic marks are not set in stone after development. They respond to the environment throughout life, and two of the best-documented influences are diet and psychological stress.

DNA methylation relies on a supply of methyl groups, and the body generates those from dietary nutrients including folate, choline, betaine, and other B vitamins through a metabolic cycle called one-carbon metabolism.¹⁶ When those nutrients are scarce, the consequences show up quickly. In animal studies, diets deficient in choline and methionine led to widespread loss of DNA methylation within a week, including at specific cancer-related genes. Restoring adequate nutrition reversed the global methylation changes within one to two weeks. In humans, low folate intake has been linked to reduced DNA methylation in several tissues.¹⁷ The practical takeaway is that your cells need a steady supply of B vitamins and related nutrients to maintain normal methylation patterns, and shortfalls can shift the landscape surprisingly fast.

Chronic stress leaves its own epigenetic footprint. Stress hormones (glucocorticoids) trigger changes in DNA methylation and histone modifications at genes involved in the stress response, and these changes can persist long after the stressor ends, contributing to vulnerability to mood and anxiety disorders.¹⁸ In one well-studied example, chronic social defeat in animal models increased repressive histone marks at the BDNF gene (which supports brain-cell health), while antidepressant treatment countered this by adding activating marks like histone acetylation.¹⁹ This suggests that some psychiatric medications work, at least in part, by rewriting stress-related epigenetic marks.

Can Epigenetic Marks Be Inherited Across Generations

This is one of the most debated questions in the field. Because epigenetic marks are largely erased and reset in each new generation, the default expectation is that they do not pass from parent to child. And in most cases that holds true. However, some animal studies have shown that certain marks appear to resist erasure during early development, raising the possibility of transgenerational epigenetic inheritance. Whether similar escape from reprogramming occurs in humans remains unclear, with the notable exception of genomic imprinting, where specific genes carry parent-of-origin methylation marks that are deliberately maintained through fertilization.²⁰

Headlines about parents’ trauma or diet being “inherited” by their children tend to outrun the evidence. The animal studies are intriguing but involve extreme exposures (severe starvation, high-dose toxins) and carefully controlled breeding. Translating those findings to humans, who live messy, genetically diverse lives, is far from straightforward. The honest state of the science is that the mechanisms for transgenerational epigenetic inheritance exist in principle, but demonstrating them convincingly in people has proven very difficult.

Epigenetic Clocks and Aging

One of the more striking discoveries of the past decade is that DNA methylation changes so predictably with age that you can use it as a biological clock. A widely used clock, developed from roughly 8,000 samples spanning 51 healthy tissues, uses 353 specific CpG sites to estimate a person’s “methylation age.” In embryonic stem cells, this age reads near zero. In adult tissues, it tracks closely with chronological age. And in cancer, methylation age is dramatically accelerated, by an average of about 36 years across 20 different cancer types studied.²¹

These epigenetic clocks have become a highly accurate molecular readout of age in humans and other vertebrates.²² Researchers are now using the gap between a person’s methylation age and their chronological age, a measure called “age acceleration,” as a biomarker. People whose methylation age runs ahead of their calendar age tend to have higher risks for certain diseases and mortality, which makes the clock a potential early-warning system. The clock also reveals something conceptually important: aging is not just about damage accumulating randomly. It follows a partially programmed epigenetic trajectory, which raises the question of whether that trajectory could be slowed or reset.

Epigenetic Marks in Cancer

Cancer cells have thoroughly scrambled epigenetic landscapes. The pattern is not simply “too much methylation” or “too little.” Instead, cancer genomes typically show both widespread loss of methylation across large stretches of DNA and targeted gains of methylation at the regulatory regions of specific tumor-suppressor genes.²³ The global loss of methylation can activate genes that should stay silent, including those that promote cell growth or genomic instability. The localized gains of methylation, concentrated at CpG-rich regions near gene promoters, can shut down tumor suppressors that would normally keep cell division in check.²⁴

Because epigenetic changes are, at least in theory, reversible, they have become attractive targets for therapy. Two main classes of epigenetic drugs have reached the clinic. Inhibitors of DNA methyltransferases (DNMT inhibitors) work by binding to the methylation enzymes and blocking them, which causes methylation to fade as cells divide and can reactivate silenced tumor-suppressor genes. The FDA has approved several of these drugs for blood cancers.²⁵ HDAC inhibitors, meanwhile, block the enzymes that remove acetyl groups from histones, forcing chromatin into a more open state. This can trigger a cascade of anti-tumor effects including cell-cycle arrest, activation of tumor-suppressor pathways, and enhanced immune recognition of cancer cells.²⁶

The challenge with these drugs is specificity. Current DNMT inhibitors and HDAC inhibitors affect the genome broadly, which means they change epigenetic marks at thousands of sites, not just the handful that are misbehaving. This can cause side effects and limits how aggressively they can be dosed. Still, they have proven effective enough to become standard treatments for certain leukemias and lymphomas, often in combination with other therapies.

Precision Epigenome Editing

The blunt-instrument problem of existing drugs is driving interest in more precise tools. Researchers have adapted the CRISPR gene-editing system to rewrite epigenetic marks at a single chosen location in the genome, without cutting the DNA itself. The approach uses a deactivated version of the Cas9 protein (called dCas9) fused to an epigenetic enzyme. A guide RNA directs the complex to a specific gene, and the attached enzyme adds or removes a mark right there.

In one proof-of-concept study, scientists used a dCas9 system fused to the active part of a TET1 demethylation enzyme to strip methyl groups from a specific gene promoter in mouse liver cells and in live mice. The targeted demethylation successfully increased the gene’s responsiveness to its normal activating signals.²⁷ These CRISPR-based epigenome editing systems have advanced rapidly and are now widely used in research to test what individual epigenetic marks actually do at specific genes.²⁸

Off-target effects remain a concern. When a dCas9-methylation tool is aimed at one spot, it sometimes modifies nearby or unrelated sites as well. Ongoing profiling studies are mapping these off-target changes to understand how tightly the editing can be controlled. If the precision improves enough, the long-term vision is to use epigenome editing therapeutically: silencing a disease-causing gene or reactivating a protective one without permanently altering the DNA sequence. Because epigenetic changes are inherently reversible, this approach could, in principle, be less risky than conventional gene editing.

Reading the Epigenome

Knowing where epigenetic marks sit across the genome is essential for both research and clinical applications, and several technologies make this possible. Whole-genome bisulfite sequencing treats DNA with a chemical that converts unmethylated cytosines but leaves methylated ones intact, allowing researchers to map every methylated CpG site at single-base resolution. Studies using this technique across multiple human tissues have identified thousands of CpG-rich regions and shown that many are linked to tissue-specific gene regulation.²⁹

Histone marks are typically mapped using a technique called ChIP-seq, which uses antibodies to pull down DNA fragments bound to a particular histone modification, then sequences those fragments to see where in the genome they came from. Combining ChIP-seq for histone marks with bisulfite sequencing for DNA methylation, RNA sequencing for gene expression, and open-chromatin assays gives a layered picture of the epigenomic landscape.³⁰ These multi-layered maps are becoming standard in both basic research and clinical studies, allowing scientists to compare the epigenomes of healthy tissue versus diseased tissue and pinpoint which marks have gone wrong.

Why Chromatin Structure Itself Matters

All of the marks and modifications discussed above converge on one physical reality: the three-dimensional packaging of DNA in the nucleus. Chromatin exists on a spectrum from tightly packed (heterochromatin) to loosely packed (euchromatin). Heterochromatin is generally gene-poor and silent; euchromatin is gene-rich and active. This packaging is maintained by networks of molecular complexes that read, write, and enforce the various epigenetic marks.³¹

Aging disrupts this organization. As cells get older, they gradually lose histones and the balance between activating and repressive histone marks shifts, leading to a blurring of the boundary between heterochromatin and euchromatin. Condensed, silent regions relax, and genes that should be kept quiet can leak through.³² This erosion of chromatin architecture is thought to contribute to the cellular dysfunction that comes with aging, connecting the physics of DNA packaging to the biology of growing old. It also helps explain why the epigenetic clocks described earlier work so well: they are tracking the downstream methylation consequences of a deeper structural unraveling.