A start codon is a codon in messenger RNA (mRNA) at which a ribosome begins translation, the process of synthesizing a protein from an RNA template. The most common start codon is AUG, which specifies methionine in the standard genetic code. Other codons can initiate translation in particular organisms or molecular contexts. Recognition of a start codon depends on the surrounding RNA sequence and the translation machinery, not simply on the presence of a particular triplet.
Sequence and reading frame
Codons consist of three nucleotides, conventionally written in the RNA’s -to- direction. AUG contains adenine, uracil, and guanine. In the corresponding coding strand of DNA, the sequence is written ATG because DNA contains thymine instead of uracil.
Selection of a start codon establishes the reading frame: the grouping of successive nucleotides into triplets. For example, the RNA sequence
can be translated as methionine–alanine–lysine, followed by termination at UGA. The initiating AUG supplies the first amino acid; it is not merely a marker preceding the translated sequence.
A start codon is functionally distinct from an AUG encountered during elongation. An internal AUG normally adds methionine to an existing protein chain without initiating a new chain. Likewise, not every AUG in an mRNA is used as a start site.
An open reading frame is a sequence that can be read in a particular frame without encountering an in-frame stop codon. Identifying a plausible start and an open reading frame is useful for predicting protein-coding regions, but does not by itself demonstrate that translation occurs.
Recognition by the translation machinery
Initiation requires an initiator transfer RNA (tRNA), which is specialized for beginning protein synthesis. Its anticodon pairs with the start codon while initiation factors help assemble and position the ribosome.
The initiator tRNA occupies the ribosomal P site, unlike incoming tRNAs during elongation, which normally enter the A site. After correct initiation-complex assembly, the ribosome proceeds along the mRNA, adding amino acids through peptide bonds.
The identity of the initiating amino acid depends chiefly on the initiator tRNA and the biological system:
- In the cytoplasm of eukaryotic cells and in archaea, initiation normally uses methionine.
- In bacteria, initiation normally uses N-formylmethionine, a chemically modified methionine.
- Translation in mitochondria and chloroplasts has organelle-specific features and commonly uses formylated initiator methionine.
The initial methionine is not necessarily retained in the mature protein. Enzymatic processing can remove its formyl group, the methionine itself, or a longer amino-terminal segment.
Initiation in bacteria
Many bacterial mRNAs contain a Shine–Dalgarno sequence upstream of the start codon. This sequence can pair with a complementary region of the small subunit’s ribosomal RNA, helping position the start codon for recognition.
Start-site selection also depends on the distance between this ribosome-binding sequence and the start codon, the local RNA structure, and initiation factors. A Shine–Dalgarno sequence is not universal: some bacterial mRNAs lack a recognizable one, and leaderless mRNAs, which begin at or very near their start codon, use different initiation arrangements.
AUG is the predominant bacterial start codon. GUG and UUG are also established alternatives; additional codons can function as starts less frequently. At an initiation site, GUG or UUG generally recruits the initiator tRNA and therefore introduces formylmethionine, rather than the valine or leucine those codons normally specify during elongation.
Bacterial mRNAs may encode several proteins. Each coding region can have its own initiation site, although translation of neighboring regions may be coupled.
Initiation in eukaryotes
In the common cap-dependent scanning pathway, the small ribosomal subunit, initiation factors, and initiator tRNA are recruited near the mRNA’s cap. The complex then scans along the RNA until it recognizes a suitable start codon.
Recognition is influenced by the Kozak sequence, the nucleotide context surrounding the start site. In vertebrates, a purine three nucleotides before the A of AUG and a G immediately after AUG are particularly favorable features. These preferences are not absolute rules, and their details differ among eukaryotic groups.
The first AUG is therefore not invariably selected. Several processes can alter initiation:
- Leaky scanning: a scanning complex bypasses a start codon in an unfavorable context and initiates farther downstream.
- Reinitiation: after translating a short upstream coding region, a ribosomal complex can resume scanning and initiate again.
- Internal initiation: some RNAs recruit translation machinery through internal RNA elements rather than exclusively through the usual cap-dependent route.
- Non-AUG initiation: near-cognate codons, such as CUG, GUG, or UUG, can initiate translation at particular sites.
Non-AUG initiation is generally less efficient than initiation at a favorable AUG, but it can have important regulatory functions. Initiation at a non-AUG codon often still incorporates methionine; specialized exceptions also exist.
Regulation and alternative start sites
Start-site selection is an important level of gene expression control. RNA structures or bound regulatory proteins can expose or conceal an initiation region, changing how efficiently a protein is produced without altering its amino acid sequence.
An upstream open reading frame is a translated region preceding a main protein-coding region. Its translation can influence whether ribosomes reach and initiate at the downstream start codon. Such arrangements allow translation to respond to cellular conditions, including nutrient availability and stress.
Alternative initiation sites within the same mRNA can produce proteins with different amino termini. These differences may affect protein stability, activity, or localization. Changes in transcription or alternative splicing can also alter which start sites are present or accessible in a mature mRNA.
A mutation that removes a normal start codon may reduce translation, prevent production of the usual protein, or redirect initiation to another site. Conversely, a mutation can introduce a new initiation site. Its effect depends on sequence context and on which alternative sites the translation machinery can use.
Identification and experimental interpretation
Start codons are central to genome annotation, but their identification cannot always be reduced to finding the first ATG in a DNA sequence. Predictions must account for the organism’s initiation mechanisms, alternative start codons, transcript structure, and evidence that the proposed region is translated.
Ribosome profiling, targeted sequence changes, reporter assays, and proteomics can provide evidence for initiation sites. Each has limitations: a sequence capable of initiating translation in an artificial reporter may not be used in its native setting, and a ribosome-associated RNA segment does not automatically establish the production of a stable protein.
A start codon must also be distinguished from a promoter and a transcription start site. A promoter participates in the initiation of RNA synthesis; a transcription start site is the first nucleotide transcribed into RNA. A translation start codon marks where the RNA is decoded into a protein. These positions belong to different stages of gene expression and need not coincide.