Ink LearningBiologyPracticalsExam boards

Biology › Gene regulation, genomics and biotechnology › Controlling gene expression: which genes a cell actually reads

Controlling gene expression: which genes a cell actually reads

A liver cell and a neurone carry the same twenty thousand genes and behave nothing alike, because a gene that is never transcribed might as well not be there. Control happens at the promoter, at the chromatin, and even after the mRNA has been made — and a bacterium digesting lactose shows the whole logic in one page.

Before this Transcription and the promoter region · RNA polymerase and the template strand · Protein tertiary structure and binding sites · Cell specialisation and stem cells

Before you start

Every cell in your body contains the same DNA, so what makes a liver cell different from a neurone is that each one keeps the genes it needs and throws the rest away. The first half of that is true and the second half is the misconception the whole unit is built on dismantling. Nothing is thrown away. The nucleus of a neurone contains the gene for insulin, complete and undamaged, and the nucleus of a pancreatic β cell contains every gene a neurone uses. What differs is which of those genes is being transcribed. The evidence is not subtle: take the nucleus out of an adult mammary gland cell, put it into an egg cell whose own nucleus has been removed, and you get a whole sheep. Dolly settled this in 1996, and induced pluripotent stem cells settled it again in 2006 with nothing more than four transcription factors.

What you should be able to do

One genome, two hundred cell types

A human body runs on something like twenty thousand protein-coding genes and contains over two hundred recognisably different kinds of cell. Every one of those cells, apart from mature red blood cells and the gametes, carries the whole genome, and transcribes perhaps half of it at any moment. The half it chooses is most of what makes it the cell it is.

Gene expression can be controlled at every stage between a gene and its protein: transcription blocked or accelerated, pre-mRNA spliced one way rather than another, mRNA destroyed before a ribosome reaches it, the finished polypeptide left inactive until it is wanted. Most of the control, and almost all of the control you will be asked about, happens at the first of those, because stopping a process before it starts is cheaper than stopping it halfway.

Gene expression
The process by which the information in a gene is used to produce a functional product, usually a polypeptide.
Transcription factor
A protein that binds to a specific base sequence in DNA, usually in or near a promoter region, and so increases or decreases the rate of transcription of that gene.
Promoter
A region of DNA at the start of a gene to which RNA polymerase and transcription factors bind.

Notice what a transcription factor is made of. It is a protein, so it is the product of a gene, so it is itself under the control of transcription factors. Regulation in a eukaryote is layered several deep, and one signal arriving at the cell surface can end up changing the transcription of hundreds of genes.

Transcription factors: the general mechanism

The mechanism is the same in every case, and worth learning as a sequence rather than as a definition. A transcription factor has a region whose tertiary structure is complementary to a specific short sequence of bases, usually in or beside the promoter of the gene it controls. It binds there by hydrogen bonds and ionic interactions with the exposed edges of the bases. Once bound, it either helps RNA polymerase attach and start transcribing, or gets in the way so that RNA polymerase cannot. Factors that increase the rate are called activators, those that decrease it repressors, and both are transcription factors — so an answer saying only that a transcription factor 'switches genes on' will lose marks the moment a question hands you a repressor.

The specificity is the specificity you met in enzymes: a site with a particular shape and charge fits one short stretch of DNA in three billion base pairs well enough for binding to last. Change the factor's shape and the fit is destroyed, or suddenly created. Every example below is a variation on that.

The lac operon: how a bacterium decides

Escherichia coli prefers glucose, but can manage on lactose because it has genes for the enzymes that bring lactose in and split it. Making those enzymes when there is no lactose about would waste amino acids and ATP, so the genes stay off until lactose appears. The arrangement that manages it, worked out by François Jacob and Jacques Monod in 1961, is the clearest control circuit in biology.

An operon is a group of structural genes transcribed together as a single mRNA, plus the control sequences that switch them. The lac operon has three structural genes: lacZ, which codes for β-galactosidase, the enzyme that hydrolyses lactose into glucose and galactose; lacY, which codes for lactose permease, the transport protein that lets lactose into the cell; and lacA, a transacetylase whose job is still argued over. In front of them sit the promoter, where RNA polymerase binds, and the operator, a short sequence between the promoter and the genes.

Watch the operator. Everything that happens in this figure happens because a protein is either sitting on that short stretch of DNA or not sitting on it, and the only thing that decides which is whether there is any lactose in the cell.

Nearby, and transcribed steadily whatever else is going on, is the regulatory gene lacI. Its product is the lac repressor, a protein with two binding sites: one shaped to fit the operator, and one shaped to fit lactose. With no lactose in the cell, the repressor binds the operator, physically obstructing the promoter. RNA polymerase cannot move past, so the three structural genes are transcribed only at a very low background rate.

When lactose enters, some of it binds to the repressor's second site. That changes the tertiary structure of the protein, including the shape of the site that fitted the operator, which no longer fits. The repressor falls off. RNA polymerase now transcribes lacZ, lacY and lacA as one long mRNA, ribosomes translate it, and within a couple of minutes the cell has all three enzymes. When the lactose has been digested there is nothing left to hold the repressor in its altered shape, it reverts, it rebinds the operator, and the tap closes.

Two points of precision. What actually binds the repressor is allolactose, an isomer of lactose made by β-galactosidase from the trickle of enzyme the repressed operon still produces — every board accepts 'lactose', but the detail explains why the operator has to be leaky. And the repressor does not stop RNA polymerase binding the promoter; it stops it moving on from there.

A board note: the lac operon is a named requirement in OCR A and CAIE 9700, and is not named by AQA — though it remains the cleanest illustration of a repressor acting as a transcription factor, which AQA does ask about.

Two mutations, two very different bacteria

Two mutant strains of E. coli are isolated. In strain 1 a mutation in lacI produces a repressor protein whose lactose-binding site no longer works. In strain 2 a mutation in the operator sequence means the repressor can no longer bind to it. Predict the behaviour of each strain in a medium containing lactose and in a medium containing none.

Strain 1's repressor still binds the operator, because that site is untouched, but lactose can no longer bind it, so nothing ever changes its shape. It stays on the operator permanently. The structural genes are never transcribed in either medium, so the bacterium grows on glucose and starves on lactose.

Strain 2 is the mirror image. Its repressor is normal but has nothing to bind to, so it never obstructs anything. The structural genes are transcribed continuously in both media. The cell survives, but wastes amino acids and ATP making enzymes for a substrate that is not there, and is out-competed by the wild type on glucose.

Oestrogen: a transcription factor you have to build first

Eukaryotes have no operons — genes for related jobs are scattered across chromosomes rather than lined up in a row — but the logic of a protein binding DNA and changing what RNA polymerase does is identical. The worked example AQA names is oestrogen.

Oestrogen is a steroid, so it is lipid-soluble, and that single property decides the whole mechanism. It diffuses straight through the phospholipid bilayer without a channel or a carrier, so it reaches the inside of the cell and then the inside of the nucleus. A peptide hormone such as insulin can do none of that: it binds a receptor on the outside of the membrane and sends its message in by a second messenger.

Follow the amber disc. It never stops being oestrogen and it never touches the DNA on its own — what binds the promoter is the complex, and the hormone's whole contribution is to change the receptor's shape so that the complex can form.

Inside a target cell, oestrogen binds to an oestrogen receptor, a protein that is itself an inactive transcription factor. Binding changes the receptor's tertiary structure and exposes its DNA-binding site. The hormone-receptor complex then binds to a specific base sequence beside the target gene — the oestrogen response element — and recruits RNA polymerase, so transcription of that gene starts or speeds up.

One detail to hold lightly: school diagrams and most mark schemes put the receptor in the cytoplasm, while in real cells the great majority of it is already inside the nucleus before any hormone arrives. Write the cytoplasmic version when a question asks for the pathway.

Only cells that make the receptor can respond to oestrogen at all, which is why the hormone reshapes breast and uterine tissue and leaves the liver alone — and why about three quarters of breast cancers are oestrogen-receptor positive and are treated with drugs such as tamoxifen, which occupies the hormone-binding site so the complex cannot form.

Oestrogen response element
A specific base sequence near a target gene to which an oestrogen-receptor complex binds, increasing the rate of transcription.
Steroid hormone
A lipid-soluble hormone that crosses the cell-surface membrane by simple diffusion and acts on an intracellular receptor.

Epigenetics, stated carefully

Two metres of DNA fit inside a nucleus six micrometres across by being wound around histones — positively charged proteins that the negatively charged phosphate groups of the backbone stick to — and then coiled again and again. DNA wound tightly onto histones cannot be reached by RNA polymerase or by transcription factors, so how tightly it is wound is a control mechanism in its own right.

Histone acetylation is the first mark. Enzymes add acetyl groups to lysine residues on the histones' protruding tails, removing the positive charge from each lysine, so the tail's attraction to the DNA's phosphate groups weakens, the chromatin loosens and the promoter becomes accessible. Transcription goes up. Histone deacetylase enzymes strip the acetyl groups off, the charge returns and the chromatin condenses. Decreased acetylation means decreased transcription: learn it in that direction, because reversing it is the commonest error in this topic.

DNA methylation is the second. A methyl group is attached to a cytosine base, usually where a cytosine is followed by a guanine. Heavy methylation of a promoter silences the gene, partly by preventing transcription factors binding and partly by attracting proteins that recruit the deacetylases — so the two mechanisms are not independent, and the chromatin ends up shut either way.

Two states of the same stretch of chromosome. The sequence is identical on both sides of the figure — count the bases if you like, there is nothing to count, because nothing has changed. What differs is what can physically get to the promoter.
Epigenetics
Heritable changes in gene expression caused by chemical modifications to DNA and to the histones associated with it, without any change to the base sequence.
Epigenome
The full set of such chemical marks on a cell's DNA and histones at a given time.
Histone acetylation
The addition of acetyl groups to histone tails, which reduces their positive charge, loosens the chromatin and increases transcription.
DNA methylation
The addition of methyl groups to cytosine bases; heavy methylation of a promoter reduces or prevents transcription of that gene.

Now the part popular accounts overstate. 'Heritable' in that definition means two different things, and the difference is enormous.

Heritable through cell division is not in doubt. When a cell replicates its DNA an enzyme copies the methylation pattern from the parent strand onto the new one, so a liver cell divides into two liver cells rather than into something undecided. That is how a differentiated cell stays differentiated through hundreds of divisions, and it is why the marks are called heritable at all.

Heritable through the germ line — passed to a child in the gamete — is a different claim. Mammalian marks are erased twice: once in the primordial germ cells as gametes form, and again in the zygote shortly after fertilisation. Both erasures are extensive, so most marks a parent acquires during life cannot reach the child. A few regions escape. Imprinted genes, where whether an allele is expressed depends on which parent it came from, are the clear case — and their marks are wiped and reset each generation according to the sex of the parent, which is the opposite of accumulating.

The studies are worth knowing and worth taking seriously. In mice a handful of loci escape erasure, and a mother's diet can change her offspring's coat colour through methylation of one of them. In humans the evidence is observational: children conceived during the Dutch Hunger Winter of 1944 to 1945 show altered methylation at some loci decades later, and records from Överkalix in northern Sweden suggest associations between grandparents' food supply and grandchildren's health. None of that demonstrates a mechanism — a fetus in a starved uterus is directly exposed, other explanations were not excluded, and no molecular route from a grandparent's diet to a grandchild's epigenome has been shown in humans.

So the sentence to write is this. Epigenetic marks are reliably passed from a cell to its daughter cells, and that is how differentiation persists. Whether experiences in one human generation routinely leave marks in the next is an open question rather than an established mechanism. When a newspaper says trauma or famine has rewritten somebody's grandchildren's genes, the marks are on top of the genes, and the inheritance is the part that is not established.

TRY IT — Reading an epigenetics claim properly

A newspaper reports a study in which male mice were trained to fear a particular smell. Their sons, who had never met their fathers and had never smelled it, startled more at that smell than control mice did. The headline reads 'Fear is written into your children's DNA'. Write two criticisms of the headline as it stands, and state what would have to be shown for the underlying claim to be established.

Check your answer

The first criticism is the wording. Nothing in the study claims the base sequence changed, and 'written into DNA' says that it did. An epigenetic mark is a chemical group attached to DNA or to the histones around it; the gene underneath is untouched, which is precisely why the effect is in principle reversible and why it should not be described as a change to the DNA.

The second is that an association between a father's experience and his sons' behaviour does not by itself identify a route. Sperm carries more than a nucleus, and a laboratory offers other paths — differences in how the sons were handled, in the mothers, or in the environment the sperm came from. The mammalian germ line also erases most marks twice, so any claimed transmission has to explain how this one survived both erasures.

To establish the claim you would want the mark identified in the father's sperm, found again in the sons' relevant tissue, shown to survive both waves of reprogramming, and shown to produce or remove the behaviour when it is added or taken away. Until then the honest summary is that something was transmitted and the mechanism is unknown.

RNA interference: switching off after transcription

Everything so far has acted before or during transcription. There is also a mechanism that lets a cell destroy an mRNA that has already been made, which is the fastest way to stop production of a protein without touching the gene at all.

Double-stranded RNA in the cytoplasm — from a virus, from a transcribed repeat, or from a laboratory — is cut by an enzyme into fragments about twenty-one nucleotides long: small interfering RNA, or siRNA. One strand of each fragment is loaded into a protein complex and the other discarded. The complex then uses its single strand as a search template, and wherever it meets an mRNA whose bases are complementary it pairs with it and cuts it, leaving the pieces to cytoplasmic enzymes. The gene was transcribed; no protein is made.

Cells also make their own short RNAs, microRNA, from their own genes. In animals these usually bind imperfectly and block translation rather than destroying the message, which is a dimmer switch rather than an off switch. One microRNA can regulate hundreds of different mRNAs.

siRNA
Small interfering RNA: short double-stranded RNA fragments, one strand of which guides a protein complex to a complementary mRNA so that the mRNA is broken down and cannot be translated.
RNA interference
The inhibition of gene expression by small RNA molecules binding to complementary mRNA.

Two uses. In the laboratory, siRNA is the standard way of finding out what a gene does: silence it and see what stops working. In medicine it has become a class of drug — patisiran, licensed in 2018, silences the mRNA for a faulty transthyretin protein. The attraction is specificity: a conventional drug binds a protein and will bind anything else shaped like it, while an siRNA is matched to about twenty-one bases of sequence.

Boards differ here. AQA names siRNA and asks for its effect on translation; OCR A treats it under post-transcriptional and translational control alongside alternative splicing; CAIE 9700 does not require RNA interference at all.

In the exam

Check yourself

A drug inhibits histone deacetylase enzymes in a group of cells. Predict its effect on the transcription of genes in those cells, explain your reasoning in terms of charge, and suggest why such a drug is being investigated as a cancer treatment rather than as a general tonic.

Answer

Histone deacetylases remove acetyl groups from histone tails. Inhibiting them means acetyl groups accumulate, because they are still being added and are no longer being taken off.

An acetyl group neutralises the positive charge on a lysine residue in the histone tail. With more acetyl groups, the histones carry less positive charge, so they are attracted less strongly to the negatively charged phosphate groups of the DNA backbone. The chromatin becomes less condensed, promoter regions become accessible to transcription factors and to RNA polymerase, and transcription of the affected genes increases.

The cancer application follows from what has often gone wrong in a tumour cell. Tumour suppressor genes are frequently silenced by heavy methylation of their promoters and deacetylation of the associated histones rather than by any mutation, so the sequence is intact and re-opening the chromatin can restore the protein, and with it the control on division. Several such inhibitors are licensed for particular lymphomas.

It would be a poor general tonic because the effect is indiscriminate: the drug increases acetylation across the genome, so genes that were properly silenced are turned up too, in every cell it reaches. That is where the side effects come from, and it is the general problem with treating a control mechanism rather than a gene.

Questions

Written to the command words the boards use. Try them on paper before opening a scheme: the marks go to points made, not to length.

Question 14 marks

Describe how the lac operon is switched on when lactose enters an Escherichia coli cell, naming the parts of the operon involved.

Mark scheme
  1. B1 the regulatory gene lacI is transcribed steadily whatever else is happening, and its product is the lac repressor, which binds to the operator
  2. B1 with the repressor on the operator, RNA polymerase cannot move on from the promoter, so the structural genes are transcribed only at a very low background rate
  3. B1 lactose binds to the repressor's other binding site and changes the protein's tertiary structure, so the site that fitted the operator no longer fits it
  4. B1 the repressor leaves the operator, so RNA polymerase transcribes lacZ, lacY and lacA as one mRNA and the enzymes are made

Question 24 marks

Explain why oestrogen can act on a gene inside the nucleus while a peptide hormone such as insulin cannot.

Mark scheme
  1. B1 oestrogen is a steroid and therefore lipid-soluble
  2. B1 it diffuses straight through the phospholipid bilayer without a channel or a carrier, so it reaches the inside of the cell and then of the nucleus
  3. B1 there it binds an oestrogen receptor, changing the receptor's tertiary structure and exposing its DNA-binding site, so the complex binds the oestrogen response element and transcription of the target gene increases
  4. B1 insulin is a peptide and is not lipid-soluble, so it binds a receptor on the outside of the membrane and its message is carried inwards by a second messenger instead

Question 33 marks

Explain how small interfering RNA stops a protein being produced from a gene that has already been transcribed.

Mark scheme
  1. B1 double-stranded RNA in the cytoplasm is cut into fragments about twenty-one nucleotides long, and one strand of each fragment is loaded into a protein complex
  2. B1 the complex uses that single strand as a search template and pairs with any mRNA whose bases are complementary to it
  3. B1 the mRNA is cut and broken down, so it is never translated and no polypeptide is made, even though transcription of the gene did happen

Question 43 marks

A drug occupies the hormone-binding site of the oestrogen receptor without changing the receptor's shape. Suggest the effect of this drug on transcription of an oestrogen-responsive gene, and suggest why it affects breast tissue far more than liver tissue.

Mark scheme
  1. B1 oestrogen can no longer bind, because the drug is occupying its binding site
  2. B1 the receptor's tertiary structure is therefore not altered, so no hormone-receptor complex able to bind the oestrogen response element forms and transcription of that gene is not increased
  3. B1 only cells that make the receptor can respond to oestrogen at all, and breast tissue makes it while liver tissue does not, so there is little for the drug to interfere with in the liver

Question 52 marks

State the effect that increased methylation of a promoter has on transcription of the gene, and state the effect of decreased acetylation of the histones around it.

Mark scheme
  1. B1 heavy methylation of the promoter reduces or prevents transcription of that gene
  2. B1 decreased acetylation also reduces transcription, because the histones regain positive charge and the chromatin condenses

Worth remembering

← Hardy-Weinberg: what alleles do in a population when nothing acts on them · Mutation and cancer: what goes wrong, and where it goes wrong →