Genetic code. Translation

Protein synthesis, or translation, is the last step in the expression of a gene. This topic covers the genetic code and its dialects, reading frames, the effect of point mutations, the wobble of tRNA, transfer RNA itself, aminoacyl-tRNA synthetase, the ribosome, initiation in prokaryotes and eukaryotes, elongation and its error control, polysomes, termination, the half-life and degradation of mRNA, how eukaryotic mRNA controls its own stability, how small RNAs control mRNA degradation, delayed translation in zygotes, how interferons manipulate translation, and translation as a target for antibiotics and toxins.

The synthesis of proteins, like that of nucleic acids, is template-based, and it could not be otherwise.

In this case, however, the product is very different from the template: a polypeptide is made against an RNA. Because genetic information is "translated" from the language of nucleic acids into the language of proteins, protein synthesis is called translation.

There is a further difficulty. The synthesis of DNA and RNA is based on direct recognition between nucleotides. There can be no direct recognition between amino acids and the nucleotides of the RNA template. For that reason translation is a much more complex process than replication or transcription, and most of what follows is the machinery invented to bridge that gap.

A code is any way of transferring information from one set of symbols to another. The genetic code is the way of transferring genetic information from nucleic acids to proteins.

In the genetic code, every amino acid is encoded by a combination of three nucleotides. That the unit had to be three was clear even before the code was decoded, from simple arithmetic:

• DNA has 4 types of nucleotide, so single nucleotides could encode only 4 amino acids.
• Nucleotide doublets give 4² = 16, which is not enough for 20 amino acids.
• Nucleotide triplets give 4³ = 64, which is more than enough.

The nucleotide triplets are called codons.

In a DNA or mRNA chain, sequential codons are not separated by "commas", and they do not overlap: every nucleotide is included in one codon only.

The code was decoded by a series of ingenious experiments using the bacterial enzyme polynucleotide phosphorylase, which binds nucleotides together without a template.

With it, short artificial "mRNAs" of known composition were synthesised. These were added to ribosomes for translation in vitro, and the peptides that came out were sequenced. This way scientists revealed the meaning of each of the 64 codons.

Of the 64 codons, 61 encode amino acids. The other three, UAA, UAG and UGA, are signals to stop protein synthesis. They are called stop, terminal or nonsense codons.

Most amino acids are encoded by two or more codons. This redundancy of the genetic code is called degeneracy. The synonymous codons usually differ in their third nucleotide, which is a fact worth holding on to, because it explains both the wobble and the existence of silent mutations.

The codon AUG has a double role: it is the only codon for methionine, and it is also the start signal for translation.

So a message reads:

mRNA: AUG UUU ACA ... CAC UAA
protein: Met Phe Thr ... His, then stop

The code in the standard table is known as the standard code. It was initially considered universal, and it has indeed been universal in the distant evolutionary past.

With time, however, numerous genomes, mainly small ones, accumulated minor differences in their genetic codes. All of these changes serve to decrease the number of tRNAs needed, that is, to make the translation apparatus more economic.

An incomplete list of the codes:

• the Standard Code
• the Vertebrate Mitochondrial Code
• the Yeast Mitochondrial Code
• the Mould, Protozoan and Coelenterate Mitochondrial Code, and the Mycoplasma/Spiroplasma Code
• the Invertebrate Mitochondrial Code
• the Ciliate Nuclear Code
• the Echinoderm and Flatworm Mitochondrial Code
• the Bacterial and Plant Plastid Code

Every way of dividing a DNA or RNA sequence into codons is called a reading frame. A reading frame starts with an AUG codon and ends with a stop codon.

RNA has 3 possible reading frames; DNA has 6, three for each strand.

If we take some non-coding DNA, start and stop codons will occur in it at random. After each start codon a stop codon will soon come, because roughly 1 codon in 20 is a stop codon (3 out of 64). Reading frames of this kind are called closed or blocked.

If instead we take a gene and find its true start codon, there will be no stop codon for a good length, because natural selection keeps them away. Such a reading frame is called open.

The classic analogy is an unspaced English sentence:

   ...heyhasshegottheteacupandthesugar...

1  ...hey has she got the tea cup and the sug ar...
2  ...h eyh ass heg ott het eac upa ndt hes uga r...
3  ...he yha ssh ego tth ete acu pan dth esu gar...

The three reading frames have completely different meanings, and only frame 1 is open.

A change of a single nucleotide is called a point mutation. Knowledge of the genetic code lets us predict the effect of each kind.

Substitution is the replacement of a single nucleotide.

• Because of the degeneracy of the code, many substitutions produce another codon for the same amino acid and have no effect on the protein. These are silent mutations.
• Most substitutions result in a codon for another amino acid. These are missense mutations. They can cause more or less harm, as in sickle-cell anaemia, or they can create normal polymorphism.
• A few substitutions change a codon for an amino acid into a stop codon. These are nonsense mutations. By causing premature termination they destroy the function of the polypeptide more effectively than any missense mutation does.

Deletion is the removal of a nucleotide and insertion is the addition of one. Both are frameshift mutations. After the point of mutation there will be several amino acids quite different from the old ones, and then a randomly encountered stop codon. Frameshift mutations are as harmful as nonsense mutations.

The exception is worth noting: insertion or deletion of three nucleotides causes no frameshift, only an extra or missing amino acid.

Two refinements to the picture above.

• In silent mutations, the most common change is in the third nucleotide of the codon.
• Missense mutations have a much greater chance of drastically altering the conformation of the protein if a hydrophobic amino acid is replaced by a hydrophilic one, or the other way round. Such a replacement is called non-conservative.

A frameshift can also cause immediate nonsense, when the shifted frame runs straight into a premature stop codon, or extensive missense, when it runs on for a long stretch of wrong amino acids before stopping.

Does the cell need 61 different tRNAs, one for each amino-acid codon? No. The 61 codons can be read with fewer tRNAs.

The first nucleotide of the anticodon, which stands against the third nucleotide of the codon, has some freedom. As we say, it can wobble. This allows it to bind not only the complementary nucleotide but also another one.

By wobbling, a single tRNA can recognise two codons that differ in their third letter - which is exactly the pattern in which the code is degenerate.

The function of tRNA is matching amino acids to codons.

Its molecule has four loops. Three of them are large, which gives the molecule its cloverleaf secondary structure. The 3' end carries a single-stranded CCA where the amino acid binds. Opposite it, a large loop carries the anticodon, a triplet complementary to the codon.

The three-dimensional structure is L-shaped.

Transfer RNA acts as an adaptor because amino acids have no spatial affinity for codons. But one adaptor is not enough. The tRNA recognises the codon; it does not recognise its amino acid, because all tRNAs end in the same CCA.

So a second adaptor is needed. It is an enzyme called aminoacyl-tRNA synthetase, and it is the enzyme that binds the correct amino acid to the correct tRNA.

The enzyme has binding centres for:

• the tRNA;
• the amino acid;
• ATP.

In addition, there is an editing centre.

The reaction runs in two stages.

First, the amino acid is activated, meaning its energy level is raised, using energy from ATP. The amino acid and ATP bind to the active site. The enzyme breaks a macroergic bond, releasing pyrophosphate, and binds the amino acid to the resulting AMP. The product is aminoacyl-adenylate.

Then the tRNA arrives. The enzyme binds it to the amino acid and releases AMP. The product, aminoacyl-tRNA, is now ready for the ribosome.

Note that the amino acid and the tRNA are connected by a high-energy bond. That energy is used later to create the peptide bond, which is why elongation does not need a fresh input of energy for every bond formed.

Some aminoacyl-tRNA synthetases use an editing pocket to charge their tRNAs with high accuracy.

This is hydrolytic editing: the synthetase removes its own coupling errors by hydrolysing an incorrectly attached amino acid. The correct amino acid is rejected by the editing site, so only the incorrect one is removed.

The error-correction process performed by DNA polymerase shows some similarities, but it differs in one respect: there, removal depends strongly on a mispairing with the template, whereas the synthetase's editing site judges the amino acid on its own.

The ribosome is a ribonucleoprotein complex composed of two subunits, large and small, connected by Mg²⁺.

Each subunit contains at least one molecule of rRNA and dozens of ribosomal proteins. The masses of the subunits and of the rRNAs are given in S, sedimentation units, which are not additive - which is why a 50S and a 30S subunit make a 70S ribosome rather than an 80S one.

The largest rRNA is a ribozyme responsible for forming the peptide bond. This catalytic activity is called peptidyl transferase. It is worth pausing on this: the chemical heart of protein synthesis is catalysed by RNA, not by protein.

Mitochondria have ribosomes of the prokaryotic type, a fact that matters later for antibiotic toxicity.

The two subunits associate and dissociate during each cycle of translation; they are not permanently joined.

The ribosome has three binding sites for RNA.

• One is for the mRNA template, and it is located entirely in the small subunit.
• The other two are for tRNAs and span both subunits, though mostly the large one:
  • the peptidyl site, designated P;
  • the aminoacyl site, designated A.

There is also a place for tRNA exit, shown as E.

When the two subunits are joined without mRNA and tRNA, the ribosome is inactive.

1. Proteins called initiation factors bind to the small subunit and make it detach from the large one.
2. The small subunit then binds mRNA and the first aminoacyl-tRNA, which carries methionine, forming the initiation complex.
3. Once the mRNA and initiator tRNA are bound and a start codon has been reached, the initiation factors leave and the large subunit joins. The ribosome is ready for elongation.

Two details are easy to lose and both are examinable. The initiator tRNA is special, different from the tRNA that will bring methionine during elongation. And it enters the P site, whereas every other tRNA will enter the A site. Initiation factor 2 helps it bind to the small subunit.

The order of arrival differs between the two kingdoms: in prokaryotes the mRNA comes first, while in eukaryotes the tRNA comes before it.

The Shine-Dalgarno sequence, or Shine-Dalgarno box, is a six-base consensus sequence AGGAGG close to the 5' end of prokaryotic mRNAs, generally located 6 to 7 nucleotides upstream of the start codon AUG.

It is a ribosomal binding site, and it works by base pairing: the Shine-Dalgarno sequence is complementary to a segment of the 16S rRNA of the small subunit.

Eukaryotic mRNAs lack the Shine-Dalgarno sequence. More than that, its complementary sequence was deleted from the 18S rRNA long ago, so the mechanism could not work even if the box were there.

Instead, the small ribosomal subunit recognises the 5' cap of the mRNA.

Eukaryotic initiation factors carry the prefix e, for eukaryote, and there are more of them than in prokaryotes. The most important new factor is eIF4, which recognises the mRNA cap and binds it to the small subunit. eIF4 has two subunits worth naming: the cap-binding eIF4E, and the large scaffold eIF4G, which through another protein bridges the cap to the 3' tail. The tail and the cap remain connected throughout translation, so a eukaryotic mRNA being translated is effectively a circle.

Mutations causing overexpression of eIF4E have been linked to some cases of autism.

After contact is made between the mRNA and the small ribosomal subunit, the complex crawls in the 5' to 3' direction until it reaches an AUG codon.

Only then does the first tRNA, with the first amino acid, arrive. The first tRNA always carries methionine.

The initiation factors then leave the complex, the large ribosomal subunit joins, and the ribosome is ready to start elongation.

The next aminoacyl-tRNA enters the A site, helped by an elongation factor called the transfer factor. If the anticodon matches the codon, the transfer factor leaves, letting the tRNA settle firmly in the A site.

The A and P sites lie parallel to one another. So as the anticodons of the two tRNAs come close together in the small subunit, at the opposite end the two amino acids also come close to each other, and to the peptidyl transferase active site.

The peptidyl transferase forms a peptide bond between the two amino acids, detaching the amino acid in the P site from its tRNA. The tRNA in the A site is now carrying the nascent peptide.

Next, an elongation factor called translocase forces the ribosome to move one codon downstream, in the 5' to 3' direction. The tRNA that was in the P site is pushed out of the ribosome, and the tRNA carrying the peptide, which was in the A site, is now in the P site. The A site is free and ready to welcome the next aminoacyl-tRNA.

The cycle continues, the peptide grows, and the ribosome advances. For its work the ribosome uses energy in the form of GTP.

The ribosome has several mechanisms to prevent the wrong amino acid from being incorporated.

1. Complementary codon-anticodon binding is the first filter.
2. In the case of incorrect binding, one of the elongation factors takes on an incorrect conformation, and the ribosome ejects the incorrect tRNA.
3. A wrongly loaded tRNA cannot stand in a position convenient for creating a peptide bond, so the chemistry itself refuses it.

When a ribosome moves away from the 5' end of an mRNA, another ribosome can bind there and initiate translation of its own.

In this way large "strings" of ribosomes form along a single mRNA, called polyribosomes or polysomes.

At very high magnification a polysome looks characteristic: the ribosomes are mushroom-like, attached by their small subunits to the mRNA, and the polypeptides protruding from the large subunits look like zigzags. The zigzag shape is due to the angles in the peptide bond.

Since prokaryotic mRNA does not mature, it forms a polysome and initiates translation while it is still being synthesised. Transcription and translation happen on the same molecule at the same time.

In an mRNA transcribed from an operon, each gene has its own Shine-Dalgarno sequence, so ribosomes can start from the middle of the molecule rather than only at its 5' end.

Genes organised into operons exist in eukaryotes too, but they are rare.

Termination begins when the ribosome reaches a stop codon.

No tRNA can bind a stop codon. Instead it is recognised by proteins called release factors. They change the activity of the peptidyl transferase, forcing it to hydrolyse the bond between the last tRNA and its amino acid, rather than transfer it to a new amino acid.

The finished polypeptide is released, and then the entire complex disassembles. The components are recycled.

mRNA turnover is defined by the mRNA half-life, the length of time an mRNA exists in the cell before it is degraded.

In eukaryotes, transcripts have a wide range of half-lives:

• Housekeeping genes such as actin and β-globin have very slow decay rates, on the order of days.
• Several proto-oncogenes such as c-Myc, and mRNAs with regulatory roles such as the cytokines, have half-lives of less than 1 hour.

Any change in the half-life of these short-lived mRNAs can dramatically alter their abundance, leading to serious changes in gene regulatory networks, cell growth and differentiation. Deregulation of mRNA stability plays an important role in the development of diseases including cystic fibrosis, muscular dystrophy and cancer.

How degradation happens. In some cases degradation of the cytoplasmic mRNA is initiated by shortening of the poly(A) tail, called deadenylation. The 5' cap is then removed by the decapping complex, and the decapping reaction promotes degradation by cytoplasmic exoribonucleases. In other cases degradation starts with decapping instead; the result is the same.

A eukaryotic mRNA carries several regulatory elements:

Of these, the one to know is the ARE. It is the place where regulatory miRNAs or proteins make contact, and its function is to control the balance between the stability of the mRNA and its degradation. For that reason it is called the stability element.

In the cytoplasm, AUBPs (AU-binding proteins) bind to the ARE, whose sequence is AUUUA, in the mRNA.

The binding of AUBPs recruits either a factor which promotes decapping of the mRNA (Dcp), or a deadenylase, which removes the poly(A) tail. Exonucleases then act to degrade the mRNA.

Selective degradation of short-lived mRNAs can also be mediated by miRNAs and siRNAs.

They function as parts of a ribonucleoprotein complex called RISC, the RNA-induced silencing complex. Complementary binding of these small RNAs to the mRNA blocks its translation and targets it for degradation.

An important detail: to block translation, partial complementarity is sufficient. Full pairing is not required.

Processing bodies (P-bodies) are cytoplasmic ribonucleoprotein granules composed primarily of translationally repressed mRNAs and of proteins related to mRNA decay, which suggests a role in post-transcriptional regulation.

P-bodies exhibit the properties of liquid droplets: they are not membrane-bound compartments but separated phases within the cytoplasm.

Sometimes protein synthesis is needed at a time when transcription is difficult or impossible. In these cases mRNAs are stored in advance in the cytoplasm as ribonucleoprotein complexes and used later.

Spermatogenesis. During the last stages, a number of proteins must be synthesised while the tight packaging of DNA with protamines prevents any new transcription.

Early development is an even better example. In most animals the embryonic genome is activated only at the so-called maternal-to-zygotic transition, which happens at the midblastula stage. Until then, all protein synthesis is based on maternal mRNAs accumulated in the oocyte cytoplasm during oogenesis. Plotted over time, there is first a phase in which only the accumulated maternal mRNAs are available, and then a phase in which embryonic mRNAs are produced.

There is a classic demonstration in sea urchin embryos grown in the presence of two antibiotics.

• With actinomycin D, which stops transcription, the embryos still developed to blastulae without any new transcription at all.
• With cycloheximide, which stops translation, the zygotes did not divide even once. Some, while dying, produced fragments.

Regulation of translation is used in antiviral defence.

A virus-infected cell synthesises and secretes a protein called interferon as a warning. Interferon binds to surface receptors on neighbouring cells and triggers a "virus alert" in them. Two important enzymes are then synthesised and activated in the alerted cell.

• A protein kinase, which phosphorylates the initiation factor eIF-2, making it inactive.
• A ribonuclease, which degrades mRNAs.

Both enzymes lead to non-specific inhibition of translation. The cell stops synthesising its own proteins in order to prevent the synthesis of viral proteins - a deliberate sacrifice.

Many antibiotics and other toxins act by interfering with translation. Translation is an attractive target precisely because the prokaryotic and eukaryotic machineries differ enough for a drug to hit one and not the other.

Puromycin is the clearest illustration of the principle. Its molecule resembles the 3' end of an aminoacyl-tRNA. It enters the A site and deceives the peptidyl transferase into transferring the nascent peptide to it. From puromycin, however, synthesis cannot continue: the peptide leaves the ribosome "stillborn". In this way every molecule of puromycin destroys a protein molecule by causing premature termination.

This table is for orientation rather than memorisation.

This is where the prokaryotic character of mitochondrial ribosomes becomes a clinical problem.

Aminoglycoside antibiotics such as gentamicin, tobramycin and amikacin inhibit protein synthesis in our own mitochondria, by binding to the small ribosomal subunit because of its similarity to the bacterial one. Some mitochondrial mutations increase that similarity, and therefore the susceptibility to the drugs.

In the inner ear the mitochondria work intensively. Depriving them of proteins disrupts their functioning and leads to the production of free radicals, apoptosis of the sensory hair cells, and permanent ear damage.

Diphtheria is recognised clinically by the dirty white pseudomembrane in the throat.

Ricin is worth a paragraph of its own. It is a lectin, a carbohydrate-binding protein, produced in the seeds of the castor oil plant. It is highly toxic: a dose of purified ricin powder the size of a few grains of table salt can kill an adult human.

It works in two parts. The RTB chain binds complex carbohydrates containing galactose residues on the membrane of eukaryotic cells, which is how the toxin enters the cytoplasm. The RTA chain then breaks the glycosidic bond within the large rRNA of the 60S subunit of eukaryotic ribosomes and completely inactivates them.

• Translation needs adaptors because amino acids have no affinity for nucleotides. There are two: tRNA recognises the codon, and aminoacyl-tRNA synthetase recognises the amino acid.
• 64 codons: 61 for amino acids, 3 stops (UAA, UAG, UGA). AUG is both methionine and start. Degeneracy is mostly in the third nucleotide, which is also where the anticodon wobbles.
• Point mutations map onto the code: silent, missense, nonsense, and frameshift. Frameshift and nonsense are the damaging ones.
• Prokaryotic initiation uses the Shine-Dalgarno box pairing with 16S rRNA; eukaryotic initiation uses the 5' cap and eIF4.
• The initiator tRNA is special and enters the P site. Every other tRNA enters the A site.
• The peptidyl transferase is a ribozyme, and the energy for the peptide bond comes from the bond made earlier by the synthetase.
• mRNA lifetime is regulated, chiefly through the ARE and through miRNA/siRNA in RISC.
• Mitochondrial ribosomes are prokaryotic, which is why aminoglycosides are ototoxic.