Genetic code

The genetic code assigns RNA codons to amino acids or termination signals. Four possible nucleobases at each of three positions give $4^3=64$ codons. In the standard genetic code, 61 specify amino acids and three are stop codons; several codons can specify the same amino acid. Variants exist, so state which genetic code is being used. See Clark, Douglas and Choi: genetic code assignments and variants.

A codon is a three-nucleotide unit read in a specified reading frame during biological translation. For example, GCU specifies alanine in the standard genetic code. Its letters are read $5^{\prime }\to 3^{\prime }$.

For an original example, choose the messenger RNA segment $5^{\prime }\text{-}\mathrm{AUG}\mathrm{GCU}\mathrm{UUC}\mathrm{UGA}\text{-}{3}^{\prime }$ and stipulate initiation at the displayed AUG in a suitable eukaryotic context. Under the standard genetic code, its first three codons specify methionine, alanine and phenylalanine; the fourth is a stop codon. The resulting peptide before processing is N–methionine–alanine–phenylalanine–C. Twelve nucleotides therefore specify three amino acid residues and termination. The displayed segment omits the surrounding initiation signals; its nucleotide sequence alone is not a complete expression system.

A codon belongs to an oriented nucleotide sequence: GCU and UCG are different ordered triplets. Four possible nucleobases, adenine, cytosine, guanine and uracil, allow $4\times 4\times 4=64$ triplets because each position can be chosen independently when enumerating possible sequences. This is a counting argument, not a claim that nucleobases occur independently in real RNA. Sixty-one triplets are sense codons, meaning amino acid assignments, and UAA, UAG and UGA are stop codons in the standard genetic code.^[ref]

The useful unit is the ordered triplet together with its reading frame and genetic code, rather than three letters viewed in isolation. That combination lets DNA sequencing data support predictions about protein sequences after identifying the relevant messenger RNA. The Nirenberg and Matthaei experiment illustrates how the correspondence became experimentally accessible.

A codon used to begin a translated reading frame. AUG is common, but initiation at near-cognate codons can occur in an appropriate context.

A codon position synthesized as a nucleotide mixture, producing several codons across a library. Equal nucleotide mixture proportions do not generally produce equal amino-acid frequencies.

A library codon with any base at the first two positions and G or C at the third. Its 32 codons represent the 20 standard amino acids unequally and include one stop codon.

A library codon with any base at the first two positions and G or T at the third. Its 32 codons represent the 20 standard amino acids unequally and include one stop codon.

In the standard genetic code, UAA, UAG and UGA are stop codons: they signal termination of biological translation rather than an amino acid assignment. See Clark, Douglas and Choi: stop codons.

A stop codon supplies no additional amino acid residue. In the constructed AUG GCU UUC UGA example, replacing UGA by UAA preserves termination after the same three amino acid residues. Replacing UUC by UAA instead terminates after methionine and alanine. These are consequences of the stated genetic code, not measurements of expression efficiency.

At a standard stop codon, a release factor promotes release of the polypeptide from the ribosome. Ordinary delivery of an amino acid by transfer RNA does not supply a “stop amino acid”.^[ref]

A release factor is a protein involved in terminating biological translation and releasing the completed polypeptide at a stop codon.^[ref]

A reading frame specifies how consecutive nucleotides are grouped into codons. Starting at position 1 of AUGGCUUAA gives AUG / GCU / UAA; starting at position 2 gives UGG / CUU with AA left over. The letters are unchanged but the groups differ. A possible grouping alone does not establish an actual biological translation initiation site.

For the twelve-nucleotide example, number positions from the $5^{\prime }$ end. The three possible forward reading frames start at positions 1, 2 and 3:Parentheses enclose incomplete trailing groups. The second and third rows do not assert initiation of biological translation in a biological cell. A chosen start fixes which positions belong together; the ribosome then advances through consecutive codons rather than searching each three-letter window independently.

A sequence interpreted in a particular codon frame without an internal in-frame stop, with the boundary convention stated. Prediction of an open reading frame alone does not demonstrate biological translation or a stable functional product.

A relatively short potential coding region under an explicitly chosen length convention. Evidence of biological translation and evidence for a stable functional peptide are different.

A change of reading frame during biological translation. It need not change the underlying DNA sequence, unlike a frameshift mutation. It can be programmed by a recoding signal or occur as a translation error.

Programmed ribosomal frameshifting changes the reading frame during translation at a specific recoding signal. A slippery sequence and a downstream RNA structure can bias the event, allowing one transcript to produce different polypeptides in controlled proportions.

A frameshift mutation changes downstream grouping into codons through insertion or deletion of a number of nucleotides not divisible by three. In the constructed messenger RNA example, insert C immediately after AUG while retaining the same initiation position: $\mathrm{AUG}\mathrm{CGC}\mathrm{UUU}\mathrm{CUG}(\mathrm{A})$. The first codon is preserved, but the original downstream groups GCU, UUC and UGA disappear. The displayed fragment now has no complete standard stop codon; what happens later depends on its continuation.

The short proof is positional: every downstream nucleotide moves by $k$ positions after an insertion of length $k$. Boundaries recur every three positions, so they retain the original phase exactly when $k$ is divisible by three. An insertion of three nucleotides can still change the peptide, but it does not shift the downstream reading frame. This distinction helps interpret genetic variants without predicting a particular phenotype from grouping alone.^[ref]

Degeneracy of the genetic code means that distinct codons can specify the same amino acid. For example, UUU and UUC both specify phenylalanine in the standard genetic code.^[ref] In the original AUG GCU UUC UGA example, changing UUC to UUU preserves the predicted amino acid order. This illustrates redundancy of assignment, not equality of every possible effect on gene expression.

Degeneracy of the genetic code differs from ambiguity: multiple inputs share one output, whereas an ambiguous assignment would give a single codon more than one meaning in the same specified decoding context. The standard genetic code table has one listed assignment per triplet. Alternative genetic codes, including those used in some mitochondria, and engineered decoding contexts must be specified separately; the standard table is not a universal statement about all organelles or synthetic systems.

In 1961, Marshall Warren Nirenberg and J. Heinrich Matthaei ground washed Escherichia coli and used centrifugation to prepare a cell-free extract. Reactions at $35^{\circ }\mathrm{C}$ combined ribosomes, soluble components, ATP regeneration and phenylalanine carrying a ${}^{14}\mathrm{C}$ isotope label, with or without polyuridylic acid, an RNA containing only U. After incubation, trichloroacetic acid precipitation, washing and counting radioactive decay detected label in insoluble material. Added polyuridylic acid strongly increased incorporation; removing ribosomes or ATP regeneration suppressed it. Product characterization supported a phenylalanine polypeptide.^[ref]

The inference connects an imposed RNA template to a specific amino acid product. Yet repeated U alone cannot distinguish one-, two- or three-letter coding units: every possible group still contains only U. Neither triplet length nor the complete genetic code follows.^[ref]