Science Wiki

Genetic mutation is a change in the nucleotide sequence of the genome of an organism, a virus, or extrachromosomal DNA or RNA. Mutations result from errors in DNA replication, from spontaneous chemical changes, from damage by physical or chemical agents, and from the activity of mobile genetic elements. They are the ultimate source of new alleles and therefore of heritable genetic variation. In population genetics and evolutionary biology, mutation supplies the raw material on which natural selection and genetic drift act, and together with gene flow and genetic recombination it determines how much heritable variation a population has.

Mutation is random with respect to fitness, in the sense that the probability that a mutation occurs does not depend on whether it would be beneficial to the organism in its current environment. This was shown by the Luria-Delbrück experiment (1943) and by the replica-plating experiments of Joshua and Esther Lederberg (1952). Mutation is not random in every respect. Rates differ among sites, among types of change, among genes, and among species, and some regions of the genome mutate much faster than others. Whether the mutation rate itself can be shaped by selection, and whether it can increase under stress, is an active area of research, but there is no evidence that organisms direct mutations to the sites where they would be useful.

Mutations can be classified by the cell in which they occur, by their scale, by their molecular nature, and by their effect.

Germline and somatic mutations

• Germline mutations occur in cells that give rise to gametes, and they can be passed to offspring. They are the mutations that matter for long-term evolution.
• Somatic mutations occur in other cells of the body. They are not inherited by offspring, but they accumulate during life, they can cause cancer and other diseases, and they cause mosaicism in which different cells of one individual have different genotypes.

Point mutations

A point mutation is a change of a single nucleotide. A substitution of one purine for another purine (A and G) or of one pyrimidine for another (C and T) is a transition. A change from a purine to a pyrimidine, or the reverse, is a transversion. There are twice as many possible transversions as transitions, but transitions are generally more frequent, so the observed transition-to-transversion ratio is often about 2 for nuclear DNA in many species, and it is greater than the 0.5 expected if all changes were equally likely.

In a protein-coding region, the effect of a substitution depends on how the genetic code reads it:

• A synonymous (silent) mutation changes the codon but not the amino acid, because the code is degenerate. Synonymous changes are not always neutral, because they can affect the speed of translation, RNA splicing, mRNA stability, and folding.
• A missense mutation changes one amino acid. Its effect ranges from none to a complete loss of function, depending on the position and on the chemical similarity of the new residue.
• A nonsense mutation creates a premature stop codon. It usually truncates the protein, and the mRNA is often degraded by nonsense-mediated decay.
• A mutation can also change a start codon, a splice site, or a regulatory sequence outside the coding region, which changes how much product is made or its form.

Insertions and deletions

An indel is the insertion or deletion of one or more nucleotides. In a coding region, an indel whose length is not a multiple of three causes a frameshift mutation, which changes the reading frame downstream and typically creates a premature stop codon and a non-functional protein. An indel whose length is a multiple of three adds or removes whole amino acids and keeps the frame. Expansions of short repeats are a special class. Expansion of a trinucleotide repeat causes diseases such as Huntington's disease (CAG repeats) and fragile X syndrome (CGG repeats), and the repeats tend to expand further in successive generations.

Structural and chromosomal mutations

Larger changes include inversions, translocations, large deletions and duplications, and changes in the number of chromosomes:

• Duplications. Gene duplication is a major source of new genes. Susumu Ohno proposed in 1970 that one copy of a duplicated gene can keep its original function while the other is free to accumulate mutations. The duplicate may become a nonfunctional pseudogene, may acquire a new function (neofunctionalization), or may divide the ancestral functions with the other copy (subfunctionalization).
• Inversions and translocations. These change the order of genes without necessarily changing their number. They can reduce fertility in heterozygotes, because pairing in meiosis is disrupted, and they can contribute to reproductive isolation. Some translocations create fusion genes, as in the BCR-ABL1 fusion of the Philadelphia chromosome in chronic myeloid leukemia.
• Changes in ploidy and chromosome number. Aneuploidy is the gain or loss of individual chromosomes, as in Down syndrome (trisomy 21). Polyploidy is the multiplication of whole chromosome sets and is common in plants.
• Mobile elements. The insertion of transposable elements disrupts genes or changes their regulation. Transposable elements are a major source of mutation in many genomes.

Effects on fitness

By their effect on fitness, mutations are deleterious, neutral, or beneficial. The fitness effect depends on the environment and on the genetic background. A mutation that is beneficial in one environment can be harmful in another, as for the sickle-cell allele HbS, which protects heterozygotes against severe malaria but causes sickle cell disease in homozygotes. Mutations also interact with each other, which is called epistasis. Loss-of-function mutations reduce or remove the activity of a gene product, and gain-of-function mutations create a new or increased activity. Dominant-negative mutations produce a product that interferes with the normal product.

Spontaneous mutation

Even in the absence of mutagens, DNA is copied and maintained with some errors. The main sources are:

• Replication errors. DNA polymerases insert a wrong nucleotide at a rate of about 10⁻⁴ to 10⁻⁵ per base before correction. Proofreading by the 3′ to 5′ exonuclease and post-replicative mismatch repair reduce the final error rate to about 10⁻⁹ to 10⁻¹⁰ per base pair per cell division in many cells. Replication slippage at repeated sequences produces insertions and deletions.
• Tautomerism. The bases can exist in rare tautomeric forms that pair with the wrong partner. The original proposal of Watson and Crick was that this could cause mutations. The contribution of rare tautomers is probably smaller than that of other mechanisms, and it is still studied.
• Depurination and depyrimidination. The glycosidic bond between the base and the sugar can hydrolyze spontaneously, which leaves an abasic site. It has been estimated that a mammalian cell loses several thousand purines per day. An unrepaired abasic site is bypassed during replication, usually with the insertion of an adenine.
• Deamination. Cytosine can lose its amino group to become uracil, which pairs with adenine, so that a C:G pair can become a T:A pair after replication unless the uracil is removed by uracil-DNA glycosylase. The deamination of 5-methylcytosine produces thymine directly, which cannot be recognized as foreign. This is the reason that CpG dinucleotides are mutation hotspots in vertebrates, with C to T transitions at rates about ten times higher than at other sites.
• Oxidative damage. Reactive oxygen species that are produced by normal metabolism oxidize bases. The best-known product is 8-oxoguanine, which can pair with adenine and cause G to T transversions unless it is repaired.
• Transposition and recombination errors. Mobile elements insert into genes, and unequal crossing over between repeated sequences causes duplications and deletions.

Induced mutation

Mutagens are physical or chemical agents that increase the mutation rate:

• Ultraviolet radiation causes covalent bonds between adjacent pyrimidines on the same strand, forming cyclobutane pyrimidine dimers and (6-4) photoproducts, which distort the helix. Error-prone bypass of the dimers produces characteristic C to T and CC to TT changes, which are seen in the genomes of skin cancers.
• Ionizing radiation (X-rays, gamma rays) produces reactive oxygen species and breaks the sugar-phosphate backbone, including double-strand breaks, and causes deletions and chromosomal rearrangements. Hermann Joseph Muller showed in 1927 that X-rays cause mutations in Drosophila.
• Chemical mutagens act in several ways. Base analogs, such as 5-bromouracil, are incorporated into DNA and mispair. Alkylating agents, such as ethyl methanesulfonate (EMS) and nitrosoguanidine, add alkyl groups to bases, which changes their pairing. Deaminating agents, such as nitrous acid, convert bases. Intercalating agents, such as ethidium bromide and acridine dyes, insert between base pairs and cause frameshifts. Some compounds, such as polycyclic aromatic hydrocarbons in tobacco smoke and aflatoxin B₁, form bulky adducts after metabolic activation and cause mutations at specific sites, with a signature that is seen, for example, in TP53 mutations in lung cancers.
• Biological agents. Some viruses integrate into the genome, and the insertion can activate oncogenes or disrupt genes.

The Ames test measures the mutagenicity of a chemical by its ability to revert histidine auxotrophic mutants of Salmonella to prototrophy, usually with a liver extract that supplies metabolic activation.

Excision, Mismatch, and Double-Strand Break Repair

Cells reduce the mutation rate by several pathways of DNA repair, so that the observed mutation rate is much lower than the rate of damage. They include base excision repair (for small lesions such as uracil and 8-oxoguanine), nucleotide excision repair (for bulky lesions such as pyrimidine dimers), mismatch repair, direct reversal (for example photolyase and O⁶-methylguanine methyltransferase), homologous recombination, and non-homologous end joining. Defects in repair cause high mutation rates and disease, such as xeroderma pigmentosum (nucleotide excision repair) and Lynch syndrome (mismatch repair). Some repair pathways, such as non-homologous end joining and translesion synthesis, are themselves error-prone.

Measurement and values

Mutation rates are measured by fluctuation tests, by mutation accumulation experiments, in which lines are propagated through bottlenecks so that selection is minimized, and by sequencing parents and offspring. Rates are usually given per base pair per generation (or per cell division). Approximate values include:

The rate in humans corresponds to about 60 to 70 new single-nucleotide variants per child, and the number rises with the age of the father, because male germ cells go through many more divisions than female germ cells. The contribution of the mother also increases with age but at a lower rate. The mutation rate per genome per replication in DNA-based microbes is roughly constant at about 0.003, which is known as Drake's rule, and the rate per site varies inversely with genome size. The rate per generation in a multicellular organism is the sum over the divisions in the germline, so it also depends on generation time.

Evolution of the mutation rate

Most mutations that affect fitness are deleterious, so selection should favor lower mutation rates. The drift-barrier hypothesis proposed by Michael Lynch holds that the rate cannot be reduced beyond the point at which the selective advantage of an antimutator allele becomes smaller than the power of genetic drift, about 1/(2_N__e). This predicts that species with smaller effective populations have higher mutation rates per site, which is consistent with data from many species. Costs of fidelity, such as the time and energy that proofreading requires, also contribute. At the other extreme, mutator alleles with defects in repair can increase in frequency in asexual populations by hitchhiking with the beneficial mutations that they produce. They are common among clinical isolates of bacteria, such as in the lungs of patients with cystic fibrosis.

Population mutation rate

In population genetics, the population mutation parameter for a diploid population is:

where N_e is the effective population size and μ is the mutation rate per generation (per site, or per locus). It is the expected heterozygosity (or nucleotide diversity) at equilibrium between mutation and drift under the neutral infinite-sites model. The expected number of new mutations entering a diploid population each generation is 2_N__e μ. For example, humans have a long-term N_e of about 10,000, and a rate μ of about 1.2 × 10⁻⁸ per site, so θ is about 4 × 10⁴ × 1.2 × 10⁻⁸ ≈ 5 × 10⁻⁴ per site. This is roughly consistent with the observation that two human chromosomes differ at about 1 site per 1,000. (Observed human diversity is about 10⁻³, and the difference reflects uncertainty in the estimates of N_e and μ and the effects of demography and selection.)

Mutation-selection balance

Deleterious mutations arise repeatedly and are removed by selection. At equilibrium, called mutation-selection balance, the frequency q̂ of a deleterious allele with mutation rate μ, selection coefficient s, and dominance coefficient h is approximately:

For example, if μ = 10⁻⁶ and a recessive allele is lethal in homozygotes (s = 1), then q̂ ≈ 10⁻³. This explains why many recessive genetic diseases are maintained at low frequencies.

Distribution of fitness effects

The distribution of fitness effects (DFE) is the distribution of the selection coefficients s of new mutations. Estimates from mutation accumulation and from population genetic data show that, for new non-synonymous mutations in coding sequences in many species, most are deleterious, a substantial fraction are neutral or nearly neutral, and only a small fraction (often less than about 1 percent) are beneficial. The shape of the DFE differs among species and among genomic regions, and estimates depend on the method. The DFE of beneficial mutations is often described by an exponential distribution, which is expected from extreme value theory. Mutations of large effect are less likely to be beneficial than mutations of small effect, a point that is the basis of Fisher's geometric model.

Fixation

The fate of a new mutation depends on the product of its selection coefficient and the effective population size. A new neutral mutation in a diploid population is fixed with probability 1/(2_N_), and the rate of neutral substitution equals the mutation rate (k = μ), as described by the neutral theory of molecular evolution of Motoo Kimura. A new beneficial mutation with a small advantage s is fixed with a probability of about 2_s_ (for N_e = N and small s), so most beneficial mutations are lost by drift soon after they arise. Mutations with |_s_| much smaller than 1/(2_N__e) behave as effectively neutral, as proposed in the nearly neutral theory of Tomoko Ohta. The fixation of a beneficial mutation reduces diversity at linked sites, which is called a selective sweep.

Experimental evolution

In the E. coli long-term evolution experiment, begun by Richard Lenski in 1988, populations of Escherichia coli have been propagated for tens of thousands of generations. They have accumulated beneficial mutations, with a gradual slowing of the rate of fitness gain, and some populations have evolved elevated mutation rates. In one population, a new ability to use citrate in aerobic conditions evolved after about 31,000 generations through a sequence of mutations, which included a gene duplication.

Inherited Disorders, Oncogenesis, and Drug Resistance

• Inherited disease. Many genetic disorders result from mutations in single genes, such as cystic fibrosis, sickle cell disease, and Duchenne muscular dystrophy. Mutations in the germline that arise in the parents' gametes cause de novo disease in the child, as in many cases of achondroplasia, which results from a recurrent mutation in the FGFR3 gene.
• Cancer. Cancer arises through the accumulation of somatic mutations in oncogenes (gain of function, such as KRAS and BRAF) and tumor suppressor genes (loss of function, such as TP53 and RB1). Mutational signatures, which are patterns of base changes that reflect particular causes such as ultraviolet light, tobacco smoke, or defects in repair, can be read from tumor genomes.
• Ageing. Somatic mutations accumulate with age in tissues, and their role in ageing is under investigation.
• Antimicrobial and drug resistance. Mutations in bacteria, viruses, and cancer cells that reduce the effect of drugs are selected during treatment, which is an example of evolution in real time.

Genetic Screens and Directed Evolution

Mutations are induced in the laboratory to study gene function and to improve organisms. Forward genetic screens use mutagens, such as EMS or radiation, to produce random mutations, and the mutants are screened for a phenotype. Reverse genetics starts from a gene and creates a mutation in it, for example by site-directed mutagenesis, by transposon insertion, by RNA interference, or by genome editing with CRISPR-Cas9, which creates double-strand breaks that are repaired with errors or with a supplied template. Methods of directed evolution use rounds of random mutagenesis, such as error-prone PCR, and selection to improve enzymes. Mutation breeding has been used to produce many crop varieties with induced mutations.

Why are CpG dinucleotides mutational hotspots for C-to-T transitions in vertebrate genomes?

In vertebrate genomes, cytosines within CpG dinucleotides are frequently methylated at the 5-carbon position to form 5-methylcytosine (5mC). When unmethylated cytosine undergoes spontaneous hydrolytic deamination, it yields uracil, which is rapidly recognized as foreign and excised by uracil-DNA glycosylase. However, spontaneous deamination of 5mC produces thymine—a canonical DNA base—creating a T:G mismatch that often escapes repair before replication, fixing a C→T transition at ≈10-fold higher rates.

Why is the transition-to-transversion ratio (Ti/Tv) in nuclear genomes greater than 0.5?

Each nucleotide can undergo one possible transition (purine↔purine or pyrimidine↔pyrimidine) and two possible transversions (purine↔pyrimidine), so purely random substitution would yield a Ti/Tv ratio of 0.5. In empirical human genome datasets, the genome-wide Ti/Tv ratio is approximately 2.0–2.1 (and ≈3.0 in coding exons) because spontaneous 5-methylcytosine deamination strongly favors C→T transitions, polymerase active sites sterically reject purine:purine transversion mispairs, and transitions are more often synonymous under the degenerate genetic code.

Why does paternal age strongly increase the number of de novo germline mutations in human offspring?

In human females, oogonia complete all mitotic divisions during fetal development before arresting in meiosis I, resulting in a fixed ~22–24 cell divisions regardless of maternal age at conception. In contrast, male spermatogonial stem cells divide continuously by mitosis every ~16 days after puberty; consequently, a 20-year-old father's sperm have undergone ~150 divisions whereas a 40-year-old father's sperm have undergone ~600 divisions, adding ≈1.5–2 de novo point mutations per year of paternal age.

• Genetic variation
• DNA repair
• Neutral theory of molecular evolution
• Population genetics
• Mutation rate
• Mutagen
• Evolution