Mutation

What a mutation actually is

DNA is a long chain of chemical units. Biologists write them as four letters: A, T, C, and G. The order of the letters carries instructions, in the same way the order of letters in a sentence carries meaning.

A mutation is any change in that sequence. Sometimes one letter swaps for another. Sometimes letters are lost or added. Sometimes a whole stretch is copied, flipped, or moved somewhere else.

A human genome holds about 3 billion letters, and you carry two copies, one from each parent. Every time a cell divides, all of it has to be copied. The copying machinery is extremely accurate. It is not perfect.

The main kinds of mutation

Cells read the instructions in a gene three letters at a time. Each three-letter group, called a codon, stands for one building block of a protein. So a sentence made of three-letter words is a fair model.

Start with this sentence:

THE BIG CAT ATE THE RAT

Substitution. One letter swaps for another. THE BIG CAT ATE THE HAT One word changes. The rest is fine.

Deletion. One letter is lost. Here the I in BIG disappears, and the reader still groups letters in threes. THE BGC ATA TET HER AT Everything after the change is nonsense. This is called a frameshift.

Insertion. One letter is added. THE BIG XCA TAT ETH ERA T Another frameshift, with the same result.

Duplication. A stretch is copied. THE BIG CAT CAT ATE THE RAT The sentence is longer but still readable, and now there is a spare word that could change without breaking anything.

Real DNA has a safety margin the sentence lacks. Several codons often stand for the same building block. GAA and GAG, for example, both mean glutamic acid. So a substitution in the last letter of a codon often changes nothing at all. These are called silent mutations.

The cystic fibrosis case shows why deletions are not always frameshifts. Removing exactly three letters takes out one codon and leaves the rest of the reading frame intact.

How mutations happen

Mutations arise in several ways.

• Copying errors. When a cell divides, the copying enzymes occasionally put in the wrong letter or slip on a repeated stretch.
• Damage. Ultraviolet light can fuse neighboring letters together. X-rays can break DNA strands. Some chemicals, including several in tobacco smoke, change letters directly.
• Jumping DNA and viruses. Some stretches of DNA can copy themselves into new places. Some viruses insert their genes into their host’s DNA. About 8 percent of the human genome is left over from ancient viruses that did this to our ancestors.

Cells also have repair systems. They proofread new copies and fix most damage before it becomes permanent. So mutation is real, but cells are far from helpless. What gets through is a tiny leftover.

How often mutations happen

In 2012, Augustine Kong and colleagues at deCODE Genetics read the complete genomes of 78 Icelandic families, both parents and a child each time. They counted the letters in each child that neither parent had.

The rate came out at about 1.2 changes per 100 million letters per generation. Spread across two copies of a 3-billion-letter genome, that works out to roughly 50 to 70 new mutations in every child. You have them. So does everyone you know.

The same study found a strong effect of the father’s age. Each extra year of the father’s age added about two new mutations on average. Most new mutations come from the father, because the cells that make sperm keep dividing throughout a man’s life, and each division is another chance for a copying error. The cells that become eggs stop multiplying before a woman is even born.

Now scale that up. About 8 billion people are alive, each carrying dozens of new mutations. Simple arithmetic suggests that, among everyone alive today, almost every letter position in the human genome has been newly mutated in someone, many times over. The exceptions would be changes so damaging that an embryo cannot develop. Most of the rest do nothing. But the supply of new variation is enormous.

The rate is also not even across the genome. A C letter followed by a G mutates many times faster than average, so some spots are hotspots. Other species have different rates, and viruses with RNA genomes mutate far faster than we do.

Neutral, harmful, or helpful

Movies make mutation look instant and dramatic. Real biology is usually quieter.

Only about 1 to 2 percent of the human genome spells out proteins. A mutation that lands somewhere else often changes nothing you could measure. A silent substitution inside a gene often changes nothing either. These are neutral mutations, and they are the most common kind.

Some mutations are harmful. They break a protein that matters, like the cystic fibrosis deletion. Experiments in bacteria and yeast agree that among mutations that do change a protein’s job, harmful ones far outnumber helpful ones. Exact proportions are hard to pin down and differ between studies and species.

A few are helpful. The best human example is lactase persistence. Most mammals switch off the enzyme that digests milk sugar after weaning. In many Europeans, a single letter change about 14,000 letters away from the lactase gene keeps the gene switched on for life. It is not inside the gene at all. It sits in a control switch. Separate mutations with the same effect arose in East Africa. You can read the full story in are humans still evolving?.

Whether a mutation helps or hurts depends on:

• where in the DNA the change occurs
• what the change does to a protein or a switch
• what environment the organism lives in

That last point matters a lot. The sickle-cell substitution causes serious disease in people with two copies. One copy gives strong protection against severe malaria. So the same letter change is harmful in one setting and helpful in another.

Random with respect to need

This is the most misunderstood part of mutation. Do living things produce the mutations they need when they need them? Two classic experiments answered the question.

In 1943, Salvador Luria and Max Delbruck grew many separate cultures of the bacterium E. coli, then exposed each one to a virus that kills it. A few bacteria in some cultures survived because they were resistant.

The key was in the counts. If the virus caused resistance, every culture should have ended up with roughly similar numbers of resistant cells. Instead, most cultures had few or none, and a handful had huge numbers. That jackpot pattern is exactly what you expect if resistance mutations happen at random times while the cultures grow. A mutation that happens early leaves a big family of resistant descendants. One that happens late leaves only a few. Luria and Delbruck shared the 1969 Nobel Prize in Physiology or Medicine, with Alfred Hershey, for their work on viruses that infect bacteria.

In 1952, Joshua and Esther Lederberg made the point visible. They grew bacterial colonies on a dish, then pressed a piece of velvet onto it and printed the same pattern of colonies onto fresh dishes. This is called replica plating. When the copies were exposed to the virus, the resistant colonies grew in the same positions on every copy. So the resistance was already present on the original dish, which had never touched the virus. The Lederbergs could go back and pick out resistant bacteria without ever exposing them.

“Random” needs one clarification. It does not mean every position mutates equally often. Hotspots exist, and some bacteria raise their overall mutation rate when stressed. What has never been found is a system that aims mutations at the specific changes an organism needs.

Key idea: Mutations do not arrive because they are needed. They arise by chance, and the environment then decides which ones stick around.

The same logic explains antibiotic resistance. A bacterium does not become resistant because it wants to survive the drug. A resistant variant arose by chance, the drug killed its competitors, and its descendants took over.

Gene duplication: how new genes appear

Mutation can do more than tweak existing genes. It can also make new ones.

The usual route is duplication. A gene gets copied by mistake, so the genome now has two. One copy keeps doing the old job. The spare copy is free to change, because the organism no longer depends on it.

Your color vision is a clear case. Most mammals have two kinds of cone cells for color. Humans, apes, and Old World monkeys have three. The genes for our red-sensitive and green-sensitive light receptors, called opsins, sit side by side on the X chromosome. Their proteins are about 96 percent identical. That near match is the signature of a duplication, estimated to have happened roughly 30 to 40 million years ago in an ancestor of these primates. A few later letter changes shifted which colors each copy responds to best.

The duplication has a side effect you may know. Because the two genes are so similar and sit so close together, they sometimes line up wrongly when eggs and sperm are made. The result can be a missing or hybrid gene. This is the main cause of red-green color blindness, which affects about 8 percent of men and about 0.5 percent of women of northern European ancestry. Men are affected more because they have only one X chromosome.

Antarctic fish offer another example. Some of them make antifreeze proteins that stop ice crystals from growing in their blood. Genetic work showed that the antifreeze gene in one group of Antarctic fish grew out of a copied gene for a digestive enzyme. Arctic cod evolved a similar antifreeze on their own, from different starting DNA.

Body cells and reproductive cells

Not every mutation matters equally for evolution.

A mutation in an ordinary body cell is called somatic. A sunburn might mutate a skin cell, and that cell’s descendants might form a mole or, sometimes, a cancer. Most cancers are driven by somatic mutations that build up during a person’s life. None of these are passed to children.

A mutation in the cells that make eggs or sperm is called germline. These are the ones that can enter the next generation, and they are the ones that count for evolution. The 50 to 70 new mutations each child carries are germline mutations from a parent.

Mutation is not the same as evolution

Mutation creates variation. Evolution is a change in how common inherited variants are in a population over generations. Mutation is one ingredient, not the whole process.

An easy way to remember this:

• mutation deals new cards
• natural selection and genetic drift decide which cards stay in the deck

Most new mutations, even helpful ones, simply vanish. A new mutation starts in one individual, and that individual might have no children, or might not pass that copy on. In 1927, J. B. S. Haldane calculated that a new mutation giving a 1 percent advantage has only about a 2 percent chance of eventually spreading through a large population. Chance rules at first. Selection only gets a firm grip once a variant becomes common enough.

Without mutation, though, populations would eventually run out of new raw material. Every variant that exists today began as a mutation in some ancestor.

Common misunderstandings

• Mutation does not mean “always harmful.” Most mutations are neutral.
• Mutation does not mean “always visible.” Most have no effect you could see.
• Mutation does not mean “a planned response to need.” The Luria and Delbruck experiment showed this in 1943.
• Mutation alone does not explain adaptation. Selection does the sorting.
• Mutation is not rare. Every person is born carrying dozens of new ones.

Where to go next

To see how useful mutations get sorted from harmful ones, read natural selection. To see how chance decides the fate of new variants, read genetic drift. To see beneficial mutations spreading in real human populations, read are humans still evolving?.