Chapter 19: Epigenetic Regulation of Gene Expression

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome back to the Deep Dive.

Today we're tackling something really fascinating, almost a paradox in biology.

Think about identical twins,

same DNA, right?

Same genome.

Exactly the same sequence.

But as they grow older, they can end up quite different.

Different health issues, different traits.

How does that happen if the blueprint is identical?

Well, that's the million dollar question, isn't it?

And it forced genetics to look beyond just the sequence.

For a long time, it was all about the A's, T's, C's, and G's.

Right.

But questions like that or even, you know, why sometimes you only use the gene copy from your mom and not your dad pushed us into epigenetics, literally meaning above or on top of genetics.

Epigenetics.

Okay, let's break that down right away.

We need to be clear about the difference between the genome and this epigenome.

Absolutely.

That's the core concept you need to hold on to.

Your genome, that DNA sequence is basically fixed.

It's set from birth, stays constant your whole life.

Okay, static.

Pretty much.

But the epigenome, well, that's the layer of instructions on the genome.

Chemical tags, structural changes, and crucially, it's different in different cells.

Your brain cells have one epigenome, your liver cells another.

Ah, so it's cell specific.

Highly cell specific.

And it changes.

It adapts throughout your life based on, well, everything.

Your diet, stress, your environment.

It's dynamic.

So the genome is the hardware of the DNA sequence, but the epigenome is like the software telling the hardware what genes to run and when.

That's great analogy, exactly.

So our mission today is to look at the three main ways this software is written and maintained.

These molecular mechanisms control so much from how we develop to why we get certain diseases.

All right, let's get into mechanism number one.

This seems fundamental.

DNA methylation, actually adding something to the DNA itself.

That's right.

It's a chemical modification.

We're talking about adding a tiny molecule, a methyl group, that's CH3, onto a cytosine base in the DNA.

This creates something called 5 -methylcytosine or 5 -MelC.

And the enzymes that do this job, the writers in this case, are DNA methyl transferases,

DNMTs for short.

DNMTs, got it.

Are there different types?

Yep.

DNMT3A and 3B are generally responsible for establishing new methylation patterns, like during development.

Then there's DNMT1, which is more of a maintenance enzyme.

It copies existing methylation patterns onto new DNA strands when cells divide.

So it preserves the pattern.

And where do these DNMTs tend to put these methyl tags?

Is it random?

Oh, definitely not random.

It happens almost exclusively where a cytosine C is followed immediately by guanine.

We call that a CPG dinucleotide.

CPG?

And often these CPGs are found clustered together in specific regions, usually near the start of a gene, in what we call CPG islands.

Okay, CPG islands near gene promoters.

So functionally, what does adding that methyl group to a CPG island do?

Turn the gene on, off.

It generally silences the gene.

Think of the methyl group sticking out into the major groove of the DNA double helix.

It physically gets in the way.

Blocks things.

Exactly.

It blocks the proteins needed to start the transcription factors from binding to the DNA.

If they can't bind, the gene can't be read.

It's switched off.

So methylation in a promoter CPG island usually means gene silencing.

Unmethylated means the gene is potentially active?

That's the general rule, yes.

And the overall pattern of methylation across the genome of the methylome is incredibly specific.

The source material highlights how you can tell a bone cell from, say, a bladder cell just by looking at their methylomes.

It reflects their different jobs.

Wow.

That's specific.

It is.

And it's not just about silencing specific genes.

Methylation also happens a lot in repetitive parts of the genome.

Things like lines and signs, mostly in tightly packed regions called heterochromatin.

Why there?

It's thought to be really important for keeping the chromosome stable, preventing these repetitive bits from jumping around or causing problems.

Genome integrity.

Okay, mechanism one, methyl tags on the DNA, especially at CPG islands, controlling gene access.

What's mechanism two?

You mentioned structural changes.

Right.

This involves the proteins that DNA wraps around the histones.

Remember, DNA isn't just floating free.

It's spooled around these histone protein cores, forming structures called nucleosomes, like beads on a string.

And the tails of these histone proteins, particularly the N -terminal tails, stick out from the core.

These tails are like platforms for adding other chemical tags.

So we're decorating the histones now, not the DNA directly.

This is where those writers, erasers, and readers come in, right?

Exactly.

It's a useful framework.

Writers are enzymes that add modifications, things like acetyl groups, methyl groups, phosphate groups.

Erasers are enzymes that remove them.

And readers are proteins that recognize specific combinations of these marks and then trigger some downstream effect, like recruiting other proteins.

Let's take acetylation as an example, adding an acetyl group.

What does that do?

Acetylation is generally an activating mark.

When acetyl groups are added to histone tails, typically to lysine amino acids, it neutralizes their positive charge.

DNA is negatively charged, so this weakens the interaction.

Loosens the grip.

Precisely.

It loosens the histone's grip on the DNA, making the chromatin structure more relaxed or open.

This open configuration allows the transcription machinery to get in and access the genes.

So acetylation equals open chromatin, gene access possible.

And removing it, desethylation would do the opposite.

Right.

Desethylases, the erasers for acetylation, remove those acetyl groups.

The positive charge on the histones is restored, they bind the DNA more tightly, and the chromatin compacts into a closed confederation, gene silenced.

It's a very direct structural control.

But it's not just the acetylation, is it?

You mentioned methylation and phosphorylation, too.

This sounds like it could get complicated fast.

Oh, it gets incredibly complex.

That's the whole idea behind the histone code.

It's not just one mark.

It's the combination of different marks on different amino acids on different histone tails that determines the functional state of that chromatin region.

Can you give an example?

Sure.

There's a specific notation,

like H3K27E3.

That means histone H3, the lysine, at position 27, has three methyl groups Me3 attached.

That particular mark is strongly associated with silenced chromatin, often regions that need to be kept off long term.

H3K27E3 means silence.

And other combinations mean active transcription, or poised for activation, or something else entirely.

Exactly.

The source mentioned something staggering.

Just looking at methylation possibilities on histone H3 alone could yield something like 280 billion different combinations.

Billion.

How does this cell even interpret that?

It highlights the sheer level of sophistication, doesn't it?

This complexity allows for incredibly fine -tuned control, essential for creating all the different cell types and functions in our bodies from the same DNA blueprint.

Okay, so we have DNA methylation, and we have histone modifications forming a complex code.

What's the third major mechanism?

The third player involves RNA molecules that don't actually code for proteins, non -coding RNAs or ncRNAs.

Uh, RNA getting involved in regulation directly?

Yes.

There are short ones, like microRNAs, but for epigenetics, the long non -coding RNAs, or lncRNAs, generally over 200 nucleotides long, are really key players.

They're transcribed from DNA, but they aren't translated into proteins.

Their job is regulatory.

How do they regulate?

What do they do?

They act in several ways, often guiding other regulatory machinery.

The book describes two main models.

One is the decoy model.

Decoy.

Yeah.

The lncRNA might bind to, say, a transcription factor, effectively acting as a sponge or decoy, preventing that factor from finding and activating its actual target genes.

Okay, so it intercepts the signal.

What's the other model?

The guide model.

This is more proactive.

Here, the lncRNA acts like a scaffold or a guide, binding to specific protein complexes, maybe the DNMTs we talked about, or histone modifiers, and physically bringing them to specific locations on the genome.

So it targets the modification machinery to the right gene.

Exactly.

It ensures that the epigenetic marks are placed precisely where they're needed to either silence or activate genes.

Think of them as directing traffic for the epigenetic machinery.

Makes sense.

You need something to coordinate all this complexity.

Okay, let's see.

These three mechanisms,

DNA methylation, histone code, lncRNAs working together.

The chapter talks about monolelic expression, MAE.

Only one gene copy active.

Right.

Normally, you have two copies of most genes, one from each parent, and both are active.

But in MAE, only one allele is expressed.

A classic example is genomic imprinting.

Imprinting.

This is the parent of origin thing, right?

Precisely.

It depends on whether the gene came from the mother or the father.

During egg and sperm formation, specific genes get marked usually by methylation as being either maternal or paternal.

Then very early in embryonic development, there's a big wave of demethylation, like wiping the slate clean.

But crucially, these imprinted readings resist that erasure.

They hold on to their parental mark.

So the embryo remembers which copy came from mom and which from dad.

For certain genes, at least.

Yes.

And this leads to parent -specific gene expression.

But if this imprinting process goes wrong, if the marks are incorrect or lost, what we call an epimutation, it can cause developmental disorders.

Like the example in the text, Beckwith -Weidman syndrome.

Exactly.

BWS.

It's an overgrowth syndrome.

Normally, a growth factor gene called IGF2 is only expressed from the paternal copy.

The maternal copy is imprinted, methylated, and silenced.

In many BWS cases, due to an epigenetic error, the maternal IGF2 copy isn't silenced.

So the embryo gets a double dose of this growth factor, leading to overgrowth.

Prader -Willi and Angelman syndromes are other examples linked to imprinting problems on chromosome 15.

So imprinting relies heavily on inherited methylation patterns.

What about other kinds of M .A .E.?

There's random M .A .E.

too.

Yes.

And the classic example here is X chromosome inactivation in female mammals.

Ah, the bar body.

The bar body.

Exactly.

Females have two X chromosomes.

Males have one X and one Y.

To balance the dosage of genes on the X chromosome, female cells randomly shut down one entire X chromosome early in development.

It becomes a highly convinced inactive structure, the bar body.

And this is random.

Which X gets shut down is chance.

It's random in each cell lineage early on, yes.

And this process is beautifully orchestrated by LNC RNAs.

An LNC RNA called Zist is transcribed from the X chromosome, destined for inactivation.

Zist RNA then coats that entire chromosome, recruiting silencing complexes, histone modifiers, DNA methyl transferases to shut it down comprehensively.

Wow.

And how does the other X stay active?

There's another LNC RNA called 6, which is made from the opposite DNA strand at the same location.

6 actively represses Zist on the chromosome that will remain active, protecting it from inactivation.

It's a neat regulatory loop.

So LNC RNAs are absolutely central to X inactivation.

But the text also mentions random M .A .E.

for genes on autosomes, the non -sex chromosomes.

That sounds surprising.

It was a big discovery.

Turns out 10 to 20 percent of our autosomal genes might show this kind of random monoallelic expression.

10 to 20 percent.

That's a lot.

It is.

Like X inactivation, the choice of which allele, maternal or paternal, gets silenced seems to be random or stochastic early in development.

This creates tissues that are actually mosaics of cells expressing one allele or the other.

And can we tell which allele is off and which is on using those histone marks we discussed?

Absolutely.

Research consistently finds a specific chromatin signature.

The inactive allele tends to be marked with that silence in histone code, H3K27E3.

While the active allele carries marks associated with active transcription, like H3K36SEME3, the histone code reflects the expression state.

Incredible.

Okay.

Let's shift gears to how this applies in disease, specifically cancer.

The book emphasizes that cancer isn't just about DNA mutations anymore.

Not at all.

The view now is that cancer develops through an accumulation of both genetic and epigenetic changes.

Epi -mutations are just as important.

And in cancer, methylation seems to go wrong in two opposing ways at once.

That's the really complex part.

On one hand, cancer cells often show global hypomethylation.

That means less methylation overall across the whole genome.

What does that cause?

It can lead to chaos.

Genes that should be silent might get turned on, including oncogenes that drive cell growth.

It also contributes to genomic instability chromosomes breaking or rearranging.

Because those repetitive elements we mentioned aren't properly silenced anymore.

So widespread under -methylation is bad.

But then there's also targeted over -methylation.

Yes, selective hypermethylation.

At the same time, the genome overall is hypomethylated.

Specific CPG islands, particularly those controlling tumor suppressor genes,

become heavily methylated and thus silenced.

Ah, silencing the brakes on cell growth.

Exactly.

Genes like BRCA1, involved in DNA repair and breast cancer risk, or MLH1, another DNA repair gene, often get silenced by hypermethylation in various cancers.

It's like the cell loses its defenses.

And epi -mutations can sometimes knock out multiple suppressor genes more efficiently than trying to mutate each one individually.

But here's the hopeful part, right?

Unlike DNA mutations, which are hard to fix, these epigenetic marks are potentially reversible.

That's the crucial difference and the basis for new therapies.

Since epigenetic states can be changed, we can develop drugs that try to reset the cancer epigenome.

Like what?

Well, drugs that inhibit those DNMT enzymes, like Vidaza.

The goal is to reduce the hypermethylation silencing the tumor suppressors, allowing them to be expressed again.

Turning the brakes back on.

Right.

And similarly, there are histone decetylase, HDAC inhibitors, like Xilinza.

They block the removal of those activating acetyl marks.

The idea is to force the chromatin around tumor suppressor genes back into an open state, reactivating them.

It's about reprogramming gene expression.

That's a really promising therapeutic angle.

Okay, finally, let's connect this back to the outside world.

How does our environment, things like diet, stress, actually change our epigenome?

And can those changes be passed on?

This is maybe one of the most profound areas.

The evidence is mounting that environmental factors can definitely leave epigenetic marks.

The rat -pup nurturing study is a classic example.

Tell us about that one.

Researchers looked at rat mothers that naturally showed high or low levels of licking and grooming maternal nurturing towards their pups.

Pups raised by low -nurturing mothers grew up to be much more anxious and stressed as adults.

Okay, a behavioral difference.

But what was the mechanism?

It was epigenetic.

In the brains of the rats that received low -nurturing, the promoter region of the gene for the glucocorticoid receptor, GR, which is crucial for managing the stress response, was found to be significantly more methylated.

Higher methylation.

So less GR protein being made.

Exactly.

Less receptor meant they couldn't handle stress hormones effectively, leading to that high anxiety behavior.

It's a direct link from early life experience to an epigenetic change, methylation, to altered gene expression to a lifelong behavioral trait.

And did that get passed on?

Remarkably, yes.

Female offspring of the low -nurtured rats tended to become low -nurturing mothers themselves, and their offspring also showed the higher GR promoter methylation and stress sensitivity.

It suggests a mechanism for transmitting acquired behavioral traits across generations via epigenetics.

Wow.

Then there's the agouti mouse example with the coat color.

That seems very visual.

The avy mouse is maybe the most striking visual demonstration of what we call metastable epi -lilies.

These mice are genetically identical, but they show a huge range in coat color, from yellow to mottled to brown, pseudo agouti.

They also vary in their susceptibility to obesity and disease.

Identical genes, different outcomes.

Why?

It all comes down to the methylation state of a specific DNA element, actually a transposable element, inserted near the agouti gene promoter.

If that element is unmethylated, the agouti gene is expressed everywhere, leading to a yellow coat and obesity.

And if it's methylated?

If it's heavily methylated, the agouti gene is silenced, and the mouse has a brown coat and is lean and healthy.

The amount of methylation varies between individuals, creating that spectrum of appearances, all from the same DNA sequence.

And could the environment influence that methylation level?

Absolutely.

That was the key experiment.

When pregnant, yellow, unmethylated mothers were fed

supplemented with nutrients known to be important for methylation, like folic acid.

Vitamin B12, their offspring were much more likely to be brown and healthy.

So the mother's diet directly altered the epigenome, and thus the physical traits of her pups.

Precisely.

It showed that maternal nutrition during pregnancy could shift the epigenetic state of these metastable alleles with long -term consequences for the offspring's health.

It's a powerful example of environmental influence on the inherited epigenome.

Okay, let's try to wrap this up.

The core message seems to be that the epigenome, these layers of DNA methylation, histone modifications, guided by non -coding RNAs, acts as this crucial interface.

It sits between our fixed genome and the ever -changing environment we experience.

That's it, exactly.

It's the dynamic interpreter of our genetic potential in the context of our lives.

And understanding this is driving huge international research efforts now, like the NIH Roadmap Epigenomics Project and the International Human Epigenome Consortium, or ICC.

What's their goal?

They aim to map hundreds, even a thousand, reference epigenomes from different human cell types and tissues in different states.

By comparing healthy and diseased epigenomes, we can pinpoint the specific epigenetic changes in specific cells that contribute to complex diseases like Alzheimer's, diabetes, schizophrenia, cancer.

It's fundamental for future diagnostics and treatments.

It really reframes how we think about inheritance and disease.

Considering those agouti mice and the rat -nurturing studies,

it really makes you wonder, doesn't it?

What's the full extent of the influence our environment, our parents' environment, even our grandparents' environment, might have on our traits and health through these epigenetic mechanisms?

It certainly gives you pause to think about how your own environment right now might be subtly shaping your own epigenome.