D2.2 Gene expression

An organism’s genotype is its genetic information; its phenotype is all its observable structural and functional characteristics. Gene expression is the process by which the information in a gene has an effect on the phenotype.

For most genes this happens in three stages:

1. Transcription: the base sequence of the gene is copied into mRNA (see D1.2 Protein synthesis).
2. Translation: ribosomes use the mRNA base sequence to make a polypeptide with a specific amino acid sequence.
3. Function of the protein: the protein folds and does its job. Many proteins are enzymes, so they affect the phenotype by catalysing a particular reaction; others act as structural proteins, transporters, receptors or hormones.

Every cell in an organism has (almost) the same genes, yet a nerve cell and a liver cell look and behave very differently. The difference lies in which genes are expressed, when, and how strongly. Cells control this at several points:

• whether, and how fast, a gene is transcribed (the rest of this page, and the main point of control);
• how long the mRNA survives before it is broken down;
• how often each mRNA is translated.

Some genes code for RNA that is not translated, such as the tRNA and rRNA used in translation; expressing these genes means transcribing them.

Cells transcribe only some of their genes, and each at its own rate. This selective transcription is controlled by proteins that recognise and bind specific base sequences in the DNA.

The promoter

Every gene has a promoter, a base sequence just upstream of the gene (on the 5′ side of the coding strand). Promoters share some common sequences, but RNA polymerase cannot start transcription by itself. First, transcription factors bind the promoter; RNA polymerase then binds to them, so it is placed at the start of the gene, facing the right way to transcribe 5′ → 3′.

Transcription factors

Transcription factors are proteins with a DNA-binding site whose shape and charge fit one particular base sequence. There are many different types, each recognising a different sequence, so:

• a cell can control which genes are transcribed by controlling which transcription factors it makes;
• genes whose promoters or enhancers share a binding sequence can be switched on or off together by one transcription factor.

Enhancers, activators and repressors

Enhancers are base sequences that can be thousands of base pairs upstream or downstream of the gene. Transcription factors called activators bind them; the DNA bends so that the bound activator contacts the proteins at the promoter, and the rate of transcription increases. Other proteins, repressors, bind sequences called silencers or block activators, so the rate falls.

Signals reaching a cell work through these proteins. A steroid hormone such as testosterone, for example, binds a receptor that is itself a transcription factor (see external factors below).

Transcription is not the only control point. Once an mRNA reaches the cytoplasm, ribosomes translate it repeatedly, so a single mRNA can produce many polypeptides. Translation continues until enzymes called nucleases break the mRNA down into nucleotides. How long an mRNA survives therefore controls how much protein is made from it, and for how long.

• In human cells, an mRNA may persist for anything from a few minutes to several days before it is broken down.
• mRNAs for proteins needed only briefly, such as signalling proteins that start cell division, are typically short-lived, so production stops soon after transcription stops.
• mRNAs for proteins made in large amounts over long periods are stable. β-globin mRNA in developing red blood cells, for example, is so stable that young red blood cells keep making haemoglobin even after they have lost their nucleus.

The poly-A tail

In eukaryotes, a poly-A tail (a chain of adenine nucleotides) is added to the 3′ end of the mRNA during processing (see D1.2 Protein synthesis). After the mRNA reaches the cytoplasm, the tail is gradually shortened. Once it is short, the mRNA is translated less and is quickly degraded by nucleases. Different mRNAs lose their tails at very different rates, which is one way their lifetimes are controlled.

Cells can also target particular mRNAs: small RNA molecules called microRNAs bind complementary sequences in specific mRNAs and block their translation or speed up their breakdown.

A multicellular organism develops from a single cell, the zygote. As its cells divide by mitosis, they differentiate: they become specialized, and structures and functions appear that were not present at the start. Epigenesis is the gradual setting up of these patterns of differentiation, so that different cells of one organism come to have different structures and roles. (Historically, the word described the idea that an organism forms gradually from an unstructured egg, rather than being pre-formed in miniature.)

Differentiation happens because different cells express different genes. Some genes are activated and others are silenced, and these states are set up and held in place by chemical modifications to the DNA and to the histone proteins it is wrapped around. These modifications are called epigenetic tags.

Phenotype changes, genotype does not

The key point about epigenetic changes is that the base sequence of the DNA is not altered.

So a muscle cell and a neuron from the same person have the same genotype but different phenotypes. Epigenetic tags are also how cells remember their identity: once a stem cell has become a skin cell, its descendants stay skin cells (see B2.3 Cell specialization for stem cells and differentiation).

Three terms describe a cell’s genetic information at different stages of expression:

• Genome: the whole of the genetic information of a cell, including every gene and the non-coding sequences. With a few exceptions (such as red blood cells, which lose their nucleus, and gametes, which are haploid), all the body cells of one person have the same genome.
• Transcriptome: all the RNA molecules being transcribed in a cell, above all its mRNAs. No cell transcribes all its genes, so the transcriptome contains copies of only some of them. It differs between cell types, and in any one cell it changes over time as conditions and signals change.
• Proteome: all the proteins a cell produces. It depends on the transcriptome, because proteins are made by translating mRNA, but the amount of each protein is not simply proportional to the amount of its mRNA, because mRNA lifetime and the rate of translation are also controlled.

The human genome has about 20 000 protein-coding genes. Some, called housekeeping genes, are expressed in nearly every cell, for example genes for the enzymes of respiration. Others are expressed in only one cell type, such as the insulin gene in pancreatic β cells.

No cell expresses all of its genes, and which genes a cell expresses decides what kind of cell it becomes. Changes to the transcriptome over time allow a cell to respond to its surroundings, for example making more of a particular enzyme when its substrate becomes available.

Methylation means adding a methyl group (–CH₃) to a molecule in place of a hydrogen atom. Two kinds of methylation are important epigenetic tags.

1. Methylation of cytosine in the promoter

In the DNA of a promoter, the base most often methylated is cytosine, usually where it is followed by guanine (a CG site, also written CpG). Methylated cytosine still pairs with guanine, so the base sequence is unchanged, but the methyl groups project into the grooves of the double helix and prevent some transcription factors binding. RNA polymerase then cannot start transcription, so:

Methylation of cytosine in a promoter represses transcription and so expression of the gene downstream.

2. Methylation of histones in nucleosomes

In a nucleosome, DNA is wound around eight histone proteins (see A1.2 Nucleic acids). Each histone has a tail of amino acids that projects from the nucleosome. Amino acids in these tails, such as lysine, can be methylated. Depending on which amino acid is methylated, this can make the chromatin more or less tightly packed and so make the DNA more or less accessible to transcription factors:

• some histone methylation represses transcription (for example, on lysine 9 of histone H3, linked to condensed heterochromatin);
• other histone methylation activates transcription (for example, on lysine 4 of histone H3, found at the promoters of active genes in open euchromatin).

You do not need to know how this is achieved.

Histone tails can carry other tags too. Acetylation of histones, shown in the simulation, generally loosens nucleosome packing and activates transcription. The combination of all the tags near a gene determines whether it is expressed.

Epigenetic tags affect phenotype, and they can be passed on without any change to the nucleotide sequence of the DNA. This is epigenetic inheritance. It can happen in two ways.

To daughter cells, through mitosis

When a cell replicates its DNA before mitosis, each new double helix has one old strand and one new strand. Methyl tags on cytosines in CG sites are on both strands, so after replication the old strand still carries its tags. An enzyme recognises these half-methylated sites and adds methyl groups to the matching cytosines of the new strand. Histone tags are also largely re-established on the new nucleosomes. As a result, daughter cells inherit the same pattern of gene expression: the descendants of a liver cell remain liver cells, and changes a cell makes in response to its environment can be kept by its descendants.

To offspring, through meiosis and gametes

It was long assumed that all tags were wiped from gametes, so every zygote would start with a “blank slate”. There is now good evidence that some tags survive meiosis and gamete formation and remain in the zygote. When a phenotypic change caused by epigenetic tags is passed from parents to offspring in this way, it is called transgenerational epigenetic inheritance. It could allow the environment experienced by one generation to affect gene expression in the next.

Unlike changes to the base sequence, which happen only through random mutation, epigenetic tags can be added and removed relatively easily, so this kind of inheritance can be reversed in later generations.

The pattern of tags in a cell is not fixed. It changes during the cell’s life, partly in response to the cell’s environment, and so the environment can change gene expression in cells and whole organisms.

Air pollution and DNA methylation (required example)

Polluted air contains fine particulates (such as PM2.5), nitrogen oxides, ozone and polycyclic aromatic hydrocarbons (PAHs) from traffic, industry and burning. Studies comparing people exposed to different levels of pollution have found that:

• exposure is linked to a loss of methyl tags across the genome (lower overall DNA methylation) in blood cells and airway cells;
• the methylation of particular genes also changes; at some promoters methylation rises and at others it falls;
• the expression of genes involved in inflammation and the immune system changes, often increasing.

These changes are associated with higher rates of asthma, inflammation and cardiovascular disease in people living with polluted air. Because altered tags can be copied when cells divide, effects can persist after the exposure.

Other examples

• Diet: in agouti mice, a pregnant mother’s diet rich in methyl-group donors (such as folate) increases methylation of a gene that affects coat colour and body mass in her pups.
• Famine: people conceived during the Dutch Hunger Winter famine of 1944–45 still had altered methylation of a growth gene (IGF2) about 60 years later.

During the development of eggs and sperm in mammals, most epigenetic tags are removed, and a second wave of removal happens in the early embryo. This resets most genes, so the cells of the early embryo are totipotent. But removal is not complete:

• New tags are added to a small group of genes during egg or sperm formation, and these differ between eggs and sperm. They are not removed in the early embryo.
• So for these genes the offspring’s cells “remember” which parent each copy came from: the copy from one parent is expressed and the copy from the other is silenced or reduced. This is genomic imprinting.

One consequence is that a dominant allele can be silenced by an imprint, so a heterozygous individual may show the phenotype of the recessive allele. Another is that hybrids from reciprocal crosses (where the species of the father and mother are swapped) can differ.

Ligers and tigons

Both hybrids have one set of lion chromosomes and one set of tiger chromosomes, so they have the same autosomal genes; any difference in base sequence (such as X- and Y-linked genes in males, or mitochondrial DNA, inherited from the mother) cannot explain the size difference. The usual explanation is imprinting of genes that control growth:

• In lions, the imprinted genes from the father strongly promote growth, and the imprinted genes from the mother strongly limit growth; in a lion cub the two balance out.
• In tigers, both effects are weaker.
• A liger inherits strong growth-promoting imprints from its lion father but only weak growth-limiting imprints from its tiger mother, so growth is not held back.
• A tigon inherits weak growth-promoting imprints from its tiger father and strong growth-limiting imprints from its lion mother, so it is no larger than its parents.

There are two types of twins:

• Dizygotic (fraternal) twins develop from two eggs fertilized by two different sperm. They are no more alike genetically than other brothers and sisters.
• Monozygotic (identical) twins develop from one zygote: the early embryo splits into two parts, each of which grows into a separate individual. Apart from rare new mutations, they have the same genome.

Because monozygotic twins share their genes, any difference in gene expression or phenotype between them must be caused by the environment or by chance. This makes them a powerful way to study how the environment influences which genes are expressed.

What twin studies have shown

In one well-known study (Fraga and colleagues, 2005), DNA methylation and histone acetylation were compared in monozygotic twin pairs aged from 3 to 74 years:

• young twins (for example 3-year-olds) had almost identical patterns of tags;
• older twins (for example 50-year-olds) showed much larger differences, with different genes being expressed;
• differences were greatest in twins who had spent less of their lives together or had different lifestyles, such as diet, smoking and physical activity.

This supports the idea that epigenetic tags change during life in response to the environment, a process sometimes called epigenetic drift.

Nature of science

Twin studies are a natural experiment: the genome is a controlled variable, so environmental effects can be separated from genetic ones. But environments are not controlled, sample sizes are often small, and some differences arise by chance rather than from the environment, so results show correlation and must be interpreted carefully.

Signals from outside a cell can change its pattern of gene expression. You need one example of a hormone and one of a biochemical acting in bacteria.

A biochemical: lactose and the lac genes of E. coli

E. coli can use the sugar lactose, but only by making three proteins, coded for by three neighbouring genes (lacZ, lacY and lacA) that are transcribed together as one mRNA. lacZ codes for β-galactosidase, the enzyme that hydrolyses lactose into glucose and galactose; lacY codes for a permease that takes lactose into the cell.

• Lactose absent: a repressor protein, coded for by a separate gene (lacI), binds a base sequence called the operator, next to the promoter. RNA polymerase is blocked and the genes are hardly transcribed, so the cell does not waste energy making enzymes it cannot use.
• Lactose present: lactose that enters the cell is converted into a form (allolactose) that binds the repressor. The repressor changes shape and can no longer bind the operator, so RNA polymerase transcribes the genes and the enzymes are made.
• When the lactose has been used up, the repressor is free again and rebinds the operator, so transcription stops.

A hormone: testosterone

From puberty, Leydig cells in the testes secrete testosterone, which acts on nearby Sertoli cells to support sperm production (see D3.1 Reproduction).

1. Testosterone is a steroid, so it is lipid-soluble and diffuses through the plasma membrane.
2. In the cytoplasm it binds an androgen receptor protein.
3. The hormone–receptor complex moves into the nucleus.
4. It binds specific base sequences (androgen response elements) in the DNA and acts as a transcription factor, increasing transcription of some genes and decreasing transcription of others.