Genes and the Inheritance of Memories Across Generations

If you study architecture before having children, we do not expect your children to inherit that knowledge biologically. They might learn about architecture from the environment you create, but that is different from receiving the information through sperm or egg. Likewise, building muscle does not spare your children the need to exercise.

That distinction seems straightforward. Eye color can be inherited; experiences generally cannot. Yet research on small RNA molecules is revealing that, in some organisms, an experience can leave a biological signal that affects subsequent generations.

The important question is not simply whether memories are inherited. It is what information can cross generations, through which mechanisms, and with what consequences. Research from Dr. Oded Rechavi’s laboratory provides compelling answers in a tiny worm called Caenorhabditis elegans. What those answers mean for humans remains an open question.

Table of Contents

• DNA, RNA, and Proteins: The Instruction Book Analogy
• Why Acquired Traits Usually Stay With the Individual
• The Two Major Barriers to Inheriting Experience
• Why a Tiny Worm Is So Useful for Studying Inheritance
• RNA Interference: How Small RNAs Silence Genes
• How Worms Pass Antiviral Protection to Their Descendants
• Could a Brain Transmit Information to Future Generations?
• Neuronal Small RNAs Can Affect Behavior Across Worm Generations
• Why Germ Cells Can Influence Behavior Without Rebuilding a Memory
• Future Possibilities: Exercise, IVF, and RNA Diagnostics
• What the Evidence Does and Does Not Tell Us

DNA, RNA, and Proteins: The Instruction Book Analogy

To understand inheritance, it helps to distinguish the genetic instructions from the ways cells use them.

Think of your genome as an enormous IKEA instruction book. It contains the instructions for making all the furniture in a house, but each room needs a different selection. A kitchen does not need exactly the same things as a bedroom.

In much the same way, cells throughout the body generally contain the same genome, yet use different portions of it. A neuron and a liver cell behave differently not because each has an entirely different instruction book, but because they use different instructions.

• DNA stores the genetic instructions. Genes are regions of DNA, and the genome is the full set of genetic material.
• Chromosomes package DNA together with proteins that help organize and condense it.
• Messenger RNA carries a working copy of particular instructions, rather like a copied page for assembling one piece of furniture.
• Proteins are the products built using those messenger RNA instructions, analogous to the assembled furniture.

But messenger RNA is only one kind of RNA. Less than 2% of the human genome encodes proteins, while much more of the genome is transcribed into RNA. Some of those other RNAs regulate gene expression. For many others, we still do not fully understand their functions.

This matters because inheritance is not necessarily limited to passing along DNA sequence. RNA can carry regulatory information, including instructions that prevent particular genes from being expressed.

Why Acquired Traits Usually Stay With the Individual

One distinction is especially important: the difference between somatic cells and germ cells.

Somatic cells make up the body outside the germ line. They include muscle cells, skin cells, and the neurons involved in learning. Germ cells give rise to sperm and eggs. A sperm and an egg combine to produce a fertilized egg, from which the next individual develops.

That creates a major obstacle to inheriting experiences. When you learn something, the relevant changes occur in your nervous system. Information may be stored through changes in the strength and organization of connections between neurons. Those connections are not themselves packaged into sperm or eggs.

Even a DNA mutation acquired in a particular brain cell would not automatically appear in the DNA of the germ cells. A change occurring somewhere in the body is not, by itself, a heritable change.

Lamarck and Darwin Describe Different Processes

The classic contrast is between inheritance of acquired traits and natural selection.

In the familiar Lamarckian account, giraffes stretch their necks to reach high leaves, and the resulting increase in neck length passes to their offspring. In a natural-selection account, giraffes already differ in heritable characteristics. Those whose characteristics help them survive and reproduce in a particular environment leave more descendants.

The difference is fundamental. One explanation proposes that use or experience produces a trait that is then inherited. The other describes selection acting on existing heritable variation.

Finding an example of an acquired effect crossing generations does not mean that every experience becomes heritable, or that natural selection has been overturned. It means we need to identify the particular information being transmitted and the mechanism that allows it to travel.

The Two Major Barriers to Inheriting Experience

1. The Weismann Barrier

The separation between the soma and the germ line is often called the Weismann barrier, after August Weismann.

The central idea is that changes in the body do not automatically flow back into the reproductive cells. For an experience in the brain, gut, or muscle to influence descendants biologically, some signal would have to reach the germ line.

It is not enough to show that a parent changes and its offspring also changes. A mechanistic explanation must account for how information moves between those compartments.

2. Epigenetic Reprogramming

A second obstacle is that many molecular modifications associated with gene regulation are erased between generations.

These epigenetic modifications can influence how genetic instructions are used without changing the DNA sequence itself. In mammals, extensive reprogramming occurs in the germ line and early embryo, removing most of many classes of these marks.

Return to the instruction-book analogy. Over a lifetime, someone might underline passages, cross out instructions, or add notes in the margins. Reprogramming is like removing much of that annotation so that the next individual can begin development with access to the full instruction set.

This resetting helps explain why an acquired cellular state is not necessarily inherited. But it also sharpens the research question: could another kind of molecule carry information across generations despite these barriers?

Small RNAs are one important candidate.

Why a Tiny Worm Is So Useful for Studying Inheritance

It is difficult to overstate the importance of model organisms for understanding biology. Many fundamental insights into human cells and disease began with experiments in organisms that are easier to study and manipulate.

A model organism is supported by a research community that builds shared tools, genetic resources, and detailed biological knowledge. Because organisms share evolutionary ancestry, many cellular mechanisms discovered in one species are relevant to others. That does not make every finding transferable to humans, but it makes careful comparisons possible.

C. elegans offers an unusually powerful combination of advantages:

• A precisely mapped body and nervous system: the adult hermaphrodite has 959 somatic cells, including 302 neurons, with individually identifiable neurons and a wiring map developed beginning in the 1980s.
• Transparency: researchers can use optical tools to observe neural activity and manipulate particular cells.
• Extensive genetic resources: it was the first animal to have its genome sequenced.
• Rapid reproduction: a generation takes approximately three days, allowing researchers to follow many generations.
• Large families: one worm can produce roughly 250 offspring under the conditions discussed.
• Controlled environments: worms can be raised on plates with defined bacterial food, helping separate inherited effects from environmental differences.

Researchers in different laboratories can study the same identifiable neuron. They can also compare large numbers of genetically similar offspring across generations. Those are enormous advantages when the goal is to distinguish genetic inheritance, environmental exposure, and RNA-based effects.

In these worms, the evidence that certain acquired effects can cross generations is clear. The next task is to explain how.

RNA Interference: How Small RNAs Silence Genes

A central mechanism is RNA interference, often abbreviated RNAi.

Andrew Fire and Craig Mello showed that introducing double-stranded RNA into C. elegans could silence genes with matching sequences. Their work, published in 1998, led to the 2006 Nobel Prize in Physiology or Medicine.

Double-stranded RNA initiates a response that produces small RNA molecules. Those molecules guide the silencing machinery to matching messenger RNA. When the messenger RNA is destroyed or silenced, it cannot be used to make the corresponding protein.

In the furniture analogy, this is less like rewriting the whole instruction book and more like selectively preventing one page from being used.

Two findings made this especially relevant to inheritance:

1. Silencing could spread through the worm. Introducing the trigger into one location could produce effects elsewhere, including in germ cells.
2. Silencing could persist into the next generation. The offspring could show an effect initiated in a parent.

Researchers can also feed worms bacteria engineered to produce double-stranded RNA. The silencing response can move from the gut to other tissues and into subsequent generations. This is a routine experimental tool in worm biology, not merely an isolated observation.

However, an artificial intervention establishes what a system can do. It does not necessarily establish what the system does in response to a biologically relevant experience. That distinction motivated the next experiments.

How Worms Pass Antiviral Protection to Their Descendants

Unlike humans, these worms do not have dedicated immune cells such as T cells and B cells. One way they defend against viruses is by using small RNAs that recognize viral sequences and help silence the virus.

Rechavi and colleagues used a virus linked to a fluorescent readout. Successful viral replication made the worms glow green. When the virus was suppressed, the worms did not show that green signal.

The crucial experiment went beyond asking whether infected worms could defend themselves:

1. Parent worms encountered the virus and generated an antiviral small RNA response.
2. The researchers examined descendants lacking machinery required to initiate that response on their own.
3. They asked whether those descendants could nevertheless suppress the virus.

The descendants retained protection. The effect also continued into additional generations.

RNA sequencing identified inherited small RNAs with sequences matching the virus. Those antiviral RNAs were associated with the parents’ exposure, rather than simply being an unrelated feature of all offspring.

The inherited information was a sequence-specific molecular defense, not a change in the offspring’s DNA sequence. The parents had encountered a challenge, and their descendants received RNA-based information that helped them respond to that challenge.

This is a concrete form of biological memory across generations. It is also much narrower than inheriting a parent’s knowledge or personal recollections.

Could a Brain Transmit Information to Future Generations?

The brain makes the question particularly interesting because it integrates information about the environment and the body’s internal state.

But the brain and the reproductive system appear to use very different formats for storing information. Neural memories can involve the architecture and strength of connections among neurons. Heritable information, by contrast, must pass through the bottleneck of a fertilized egg.

How would a pattern distributed across a nervous system become a molecular signal that can travel through that bottleneck?

A Learned Response Is Not Necessarily a Transmitted Memory

Worms can learn simple associations. For example, an odor they normally approach can become less attractive when it is paired with starvation.

However, the behavioral change alone does not reveal how the learning is stored. It could involve changes in synapses. Alternatively, it might involve altered expression of an odor receptor, changing the animal’s ability to detect or respond to that odor.

That second possibility suggests a more direct molecular route. An RNA regulating a particular receptor could, in principle, influence a descendant’s response without transmitting the parent’s neural wiring.

Still, a convincing inheritance claim must explain the whole pathway. If an effect begins in the brain, how does the signal reach germ cells? If a descendant’s brain is affected, how does the inherited signal produce the appropriate change there?

Reports of inherited responses in mammals are interesting, but the discussion here does not establish that this complete chain has been convincingly demonstrated for inherited memories.

Neuronal Small RNAs Can Affect Behavior Across Worm Generations

In a 2019 study published in Cell, Rechavi’s laboratory approached the problem without requiring the conversion of a synaptic memory into a heritable molecule.

The researchers changed the production of naturally occurring small RNAs specifically in the worm’s nervous system. Descendants then showed altered food-finding behavior, even though their brains had not received the original experimental manipulation.

The effect persisted for three generations.

The mechanism involved changes in germ-line gene expression, including a gene called saeg-2. Manipulating that gene provided further evidence linking germ-line regulation to the behavioral outcome.

Importantly, the inherited effect depended on machinery involved in transmitting small RNA information across generations. Disrupting that machinery prevented the effect. Researchers could also sequence the RNAs that changed in descendants.

These experiments support a specific conclusion: altering small RNA production in a parent’s neurons can influence descendants’ behavior through an RNA-dependent inheritance pathway.

They do not show that offspring inherited a detailed memory of something a parent learned. The distinction is essential. A neural molecular state influencing descendants is remarkable enough without extending the claim beyond the evidence.

Why Germ Cells Can Influence Behavior Without Rebuilding a Memory

It may seem surprising that changing gene regulation in germ cells could alter behavior. But germ cells are not isolated from the rest of an organism’s physiology.

They can influence other tissues through chemical signals. They also contribute to the cells from which the next individual develops. An inherited signal could therefore affect behavior by changing development or physiology, rather than by traveling directly back to a particular brain circuit.

This is important for thinking about mammals as well. A small molecular difference early in development could alter later outcomes even if that original signal does not persist throughout the adult body.

The Amplification Difference Between Worms and Mammals

Worms have machinery that can amplify small RNA signals. That helps explain how an inherited effect can persist rather than simply becoming diluted across generations.

We do not know of an equivalent amplification mechanism in mammals.

That raises a central question: how could a relatively small amount of inherited RNA influence a much larger developing organism?

One possibility is timing. A signal acting when an embryo contains only a few cells might change a developmental trajectory. Another possibility is an effect on the placenta, which could subsequently influence many aspects of development.

This connects with the broader idea of the developmental origins of health and disease: events early in development can shape later physiology and health. It is a plausible framework for investigating inherited RNA effects, not proof that a particular human outcome is caused by them.

Future Possibilities: Exercise, IVF, and RNA Diagnostics

If related mechanisms are established in humans, they could eventually add a new layer to reproductive medicine. But that possibility needs to remain clearly separated from what is available now.

Could Parental Physiology Change an Inherited Signal?

Some rodent experiments suggest that parental overfeeding can produce adverse effects in offspring, and that exercise can counteract some of those effects.

Such findings raise the possibility that an inherited molecular profile might be modifiable. They do not establish a human exercise prescription for changing a future child’s inherited RNA profile.

This is different from the familiar claim that exercising builds muscles that children then inherit. The hypothesis concerns changes in biological signals associated with parental physiology, not direct inheritance of the trained body.

Could RNA Add Information to Genetic Screening?

Current reproductive genetic screening looks for relevant DNA variants. If particular RNA patterns were reliably associated with reproductive or developmental outcomes, RNA might someday provide an additional diagnostic layer.

That possibility is especially interesting because RNA profiles can be more responsive to physiological conditions than DNA sequence.

Before such an approach could be useful, researchers would need to establish which RNA patterns matter, whether they predict outcomes reliably, and whether changing them has beneficial effects.

Could RNA Be Modified During IVF?

A more speculative possibility is changing the composition of inherited RNAs during in vitro fertilization. Another is identifying a modifiable parental condition that produces a healthier RNA profile before treatment.

These remain future research possibilities, not established clinical tools. We do not yet understand human RNA-based inheritance well enough to offer them as interventions.

What the Evidence Does and Does Not Tell Us

The evidence in C. elegans establishes that certain acquired effects can pass across generations through small RNAs. Those effects include antiviral protection and changes in behavior following manipulation of neuronal RNA production.

It does not establish that humans inherit their parents’ learned knowledge, autobiographical memories, or trained abilities. Nor does it mean that every parental experience becomes biologically embedded in future generations.

The deeper lesson is that DNA sequence is not the only form of biological information worth investigating. Under particular conditions, regulatory molecules can carry information across generations and influence how descendants develop or behave.

The task now is to identify exactly which signals travel, how long they last, what they change, and where the findings apply. That is what makes this field so compelling: a fundamental assumption about inheritance is becoming a set of experimentally testable questions.