01 · Classification
Types of mutation
| Type | Change | Typical effect | Example |
|---|---|---|---|
| Substitution (transition) | A↔G or C↔T | Often synonymous (same amino acid) or missense (different AA). Nonsense if new codon is stop. | Sickle cell: A→T gives Glu→Val in HbS |
| Substitution (transversion) | Purine↔Pyrimidine (A↔C, A↔T, G↔C, G↔T) | Higher chance of missense due to different base class | Many cancer-associated mutations |
| Insertion (in-frame) | +3, +6, … bases | One or more extra amino acids inserted; protein may still function | Some Duchenne → Becker muscular dystrophy |
| Insertion (frameshift) | +1, +2, +4, +5 … bases | All downstream codons change; usually truncated or non-functional protein | Most CRISPR NHEJ outcomes |
| Deletion (in-frame) | −3, −6, … bases | One or more amino acids missing; depends on domain importance | 3-bp CFTR ΔF508 deletion |
| Deletion (frameshift) | −1, −2, −4, −5 … bases | Same as frameshift insertion — devastating downstream | Many loss-of-function disease mutations |
| Chromosomal deletion | Large segment removed | Multiple genes lost; often lethal in homozygote | DiGeorge syndrome (22q11 deletion) |
| Chromosomal duplication | Segment duplicated | Extra gene copies; may increase protein expression | Charcot-Marie-Tooth disease (PMP22 duplication) |
Substitution terminology
A substitution that changes an amino acid is a missense mutation. One that creates a premature stop codon is nonsense. One that changes the codon but not the amino acid (due to codon redundancy) is synonymous or silent. The codon table has 64 entries for 20 amino acids plus 3 stop signals — this redundancy buffers the genome against some substitutions. See Protein synthesis for the full codon table.
02 · Causes
How mutations arise
Replication errors
DNA polymerase adds the wrong base approximately once every 10⁷–10⁸ base pairs during replication. Proofreading (3′→5′ exonuclease activity) reduces this to ~1 in 10⁹. Mismatch repair lowers the final error rate to approximately 1 in 10¹⁰. For the human genome (3.2 × 10⁹ bp), this means roughly 1–2 new mutations per cell division — still enough to matter over a lifetime of ~10¹⁶ cell divisions.
Mutagens
- UV radiation (254 nm) — causes adjacent thymine bases to covalently bond (thymine dimers), distorting the helix and stalling replication. Nucleotide excision repair (NER) removes them; failure causes xeroderma pigmentosum.
- Ionising radiation (X-rays, gamma, alpha) — causes single- and double-strand breaks. Double-strand breaks are repaired by NHEJ (error-prone) or HDR (accurate but slower).
- Alkylating agents (e.g. mustard gas, some chemotherapy drugs) — add alkyl groups to bases, causing mispairing during replication.
- Base analogues (e.g. 5-bromouracil) — are incorporated in place of a normal base and pair with the wrong partner, causing substitutions.
- Intercalating agents (e.g. ethidium bromide, acridine dyes) — wedge between base pairs, causing insertions or deletions during replication.
03 · Consequences
How mutations change proteins — and phenotypes
The effect of a mutation on phenotype depends on: where in the gene it occurs, what change it makes to the amino acid sequence, and whether the affected amino acid is in a functionally critical region of the protein (active site, structural domain, binding interface).
SICKLE CELL
A → T substitution at codon 6 of beta-globin. Glu (hydrophilic) → Val (hydrophobic). Under low O₂, HbS polymerises → sickle shape → haemolytic anaemia. Heterozygotes gain malaria resistance.
CYSTIC FIBROSIS
ΔF508: 3-bp deletion removes phenylalanine 508 from CFTR. Protein misfolds, is degraded in ER, and never reaches the membrane. Results in viscous mucus in airways and pancreatic ducts.
PROGERIA
G608G in LMNA: a synonymous substitution activates a cryptic splice site, producing a truncated lamin A (progerin) that destabilises the nuclear envelope. CRISPR base editing corrects this in preclinical trials.
CANCER DRIVER
G12V in KRAS: Gly → Val locks the GTPase in active state, driving continuous cell proliferation. Present in ~25 % of all cancers. A major target for new small-molecule inhibitors.
Not all mutations affect function directly. A mutation in a regulatory region can increase or decrease gene expression without changing the protein sequence at all. Copy-number variations (CNVs) duplicate or delete entire stretches of chromosome, changing gene dosage.
04 · Evolution
Mutation as the engine of evolution
Without mutation, there would be no heritable variation; without variation, natural selection has nothing to act on. Mutations create new alleles. Natural selection retains alleles that improve survival and reproduction, removes those that reduce it, and is indifferent to neutral variants. Over many generations and populations, this process drives evolution.
The molecular clock uses the steady rate of neutral substitutions to estimate when two species diverged. Synonymous (silent) substitutions accumulate at an approximately constant rate because they are not under selective pressure — this provides a molecular timescale independent of the fossil record.
CRISPR technology is allowing researchers to correct pathogenic mutations with unprecedented precision. Base editors and prime editors can make single-letter corrections without double-strand breaks, dramatically reducing off-target effects. See CRISPR base editing for the latest clinical applications.
Mutation — common questions
- Most mutations are neutral. Synonymous substitutions change the codon but not the amino acid (due to codon redundancy). Mutations in non-coding DNA are often silent. Beneficial mutations — such as the heterozygote advantage of HbS against malaria — can be favoured by natural selection.
- Somatic mutations (e.g. in a skin cell) affect only that cell lineage. Cancer often results from accumulating somatic mutations in tumour-suppressor genes or proto-oncogenes. Germline mutations — in sperm, eggs, or early embryos — are heritable and appear in every cell of the next generation. See Genetics and inheritance.
- Replication errors are corrected by proofreading and mismatch repair — the residual error rate is ~1 in 10¹⁰ bases per replication. Mutagens increase this rate: UV causes thymine dimers; ionising radiation causes strand breaks; chemical mutagens modify bases. See the DNA guide.
- Inserting or deleting bases that are not a multiple of three shifts the reading frame — all codons after the mutation are read differently. This typically produces a premature stop codon or a drastically altered protein. CRISPR-Cas9 often introduces frameshifts via NHEJ repair — see CRISPR base editing for why base editors avoid this.
- CRISPR base editors convert one base to another (C→T or A→G) without making a double-strand break. This is especially useful for correcting point mutations that cause disease. For the clinical details — including sickle cell and progeria corrections — see CRISPR base editing.