When DNA Goes Wrong: The Science of Mutations

In 1949, a young chemist at Caltech named Linus Pauling held the page proofs of a paper barely four pages long. It carried an unassuming title, "Sickle Cell Anemia, a Molecular Disease," but that phrase would reorganize the next half century of medicine. For the first time, a human illness had been traced not to a vague imbalance of humors, nor to a microbe, but to a single defective molecule. Pauling and his colleagues had shown that the hemoglobin of people with sickle cell anemia behaved differently in an electric field from the hemoglobin of healthy people. The difference was tiny, but it was real, measurable, and physical. Somewhere in the chemistry of one protein lay the whole story of the disease.

What Pauling could not yet see was the cause of that chemical difference, because the structure of DNA would not be solved until 1953, and the genetic code would not be cracked until the 1960s. But the trail he opened led, eventually, to a single letter. Out of the roughly three billion letters that make up a human genome, the difference between health and a lifetime of painful crises came down to one. To understand how that is possible, we need to look at what a mutation actually is, how mutations come in different sizes, and why some of the most harmful ones have stubbornly refused to disappear.

What We Mean When We Say Mutation

A mutation is any change in the DNA sequence that is heritable when the cell divides. That definition is worth slowing down over, because it carries two ideas that are easy to miss. The first is that a mutation is a change in the sequence itself, the actual order of the chemical letters, not simply damage to the molecule or a temporary misreading. The second is that it must be passed on when the cell copies itself. A scratch in the DNA that the cell repairs before it divides leaves no lasting trace and is not, in the strict sense, a mutation. A change that survives into the daughter cells becomes part of that lineage forever.

Mutations come in many forms. A mutation can be a single base swapped for another, a missing or extra base, or an entire chunk of chromosome picked up and moved somewhere else. The size of the change ranges from one chemical letter to whole segments of a chromosome. Yet the most important fact about mutations is also the most counterintuitive: most of them cause nothing visible at all. The genome is large, much of it does not code for protein, and the genetic code itself has built-in redundancy. A great many changes land in places where they make no difference, or are silently corrected, or are simply tolerated. Mutation is not a rare catastrophe. It is a constant background hum, and disease emerges only when a change lands in exactly the wrong place.

Reading a Cell's Chromosomes in a Single Photograph

Before we follow mutations down to the level of single letters, it helps to zoom all the way out, to the scale at which a doctor in 1959 could actually see something go wrong. The tool for that is the karyotype, a photograph of all the chromosomes of a single cell, arranged and sorted by size. A human karyotype lays out 22 numbered pairs of autosomes plus the two sex chromosomes, 46 chromosomes in all. The technique was perfected in 1956, once researchers learned to catch cells in the middle of division, when the chromosomes are condensed and distinct, and to spread them out cleanly for the camera.

The karyotype remains the standard diagnostic test for chromosomal disorders, the kind of mutation visible at the largest scale. You cannot see a single swapped base under a microscope, any more than you could spot one misprinted letter in an entire library from across the street. But you can absolutely see that a chromosome is missing, broken, or present in triplicate. The karyotype is what makes the largest class of mutations diagnosable by eye, and it sets up a useful way of thinking about the whole subject: mutations sort themselves by scale.

The Three Scales at Which DNA Can Change

At the smallest scale sit point mutations, in which one base is swapped for another, a single letter changed in the text. One rung up are frameshift mutations, which insert or delete bases in numbers that are not multiples of three. That detail matters enormously, because the cell reads DNA in three-letter words called codons, each of which specifies one amino acid. Add or remove one or two letters, and every codon downstream of that point is regrouped into the wrong words. The reading frame shifts, and from the mutation onward the protein is gibberish. It is the difference between "the big red dog" and, after deleting one letter, "heb igr edd og." Everything after the cut becomes nonsense.

At the largest scale are chromosomal mutations, which alter the structure or the number of whole chromosomes. These are the changes a karyotype can reveal: an entire chromosome duplicated, a piece deleted, a segment flipped or moved to a different chromosome. The three scales (point, frameshift, and chromosomal) give us a clean framework. Almost every named genetic disorder is, at bottom, an example of one of them, and the rest of this article walks through the classic cases that biology students meet again and again.

When One Letter Is Enough: Point Mutations and Sickle Cell

A point mutation, the swap of a single base, can do one of three things, and the redundancy of the genetic code explains all three. Because several codons can specify the same amino acid, a base swap sometimes changes the codon to a synonym, producing exactly the same amino acid and the same protein. That is a silent mutation, and it does nothing. Other times the swap changes the codon to one that specifies a different amino acid, altering a single building block of the protein. That is a missense mutation. And sometimes the swap turns an amino-acid codon into a stop signal, cutting the protein short. That is a nonsense mutation, and it usually produces a truncated, nonfunctional protein.

Sickle cell anemia, the disease that started Pauling on this path, is the textbook missense mutation. The sickle cell allele is a single base change, an A swapped for a T, in the sixth codon of the beta-globin gene. That one change replaces the amino acid glutamic acid with valine at one position in the hemoglobin protein. Glutamic acid carries a charge and likes water; valine is hydrophobic and shuns it. That swap from a water-loving residue to a water-fearing one is enough to make hemoglobin molecules stick to one another and polymerize into long fibers when oxygen runs low, as it does in the small vessels of the body. The fibers stiffen and distort the red blood cells into the rigid, crescent shape that gives the disease its name. Those misshapen cells jam in capillaries and break apart early, producing the pain crises and anemia of the illness. One letter, one amino acid, one disease, the entire causal chain running from a single chemical substitution up to the lived experience of a patient.

When Whole Chromosomes Go Astray

Not every disorder hides at the level of single letters. Down syndrome is caused by trisomy 21, the presence of three copies of chromosome 21 instead of the usual two. There is nothing wrong with any individual gene; rather, an entire chromosome's worth of genes is present in one and a half times the normal dose, and that excess disrupts development in ways still not fully mapped. Down syndrome holds a particular place in history as the first human condition ever pinned to a specific chromosomal cause, identified by Jérôme Lejeune and colleagues in Paris in 1959, just three years after the karyotype technique matured. It was, in a sense, the chromosomal counterpart to Pauling's molecular disease: proof that you could point to a precise, visible cause for a human disorder.

Other disorders sit at intermediate scales and reveal genetics at its strangest. Huntington's disease is caused by a trinucleotide repeat expansion, in which the three-letter sequence CAG is repeated too many times in a row within the HTT gene on chromosome 4. Normal alleles carry between 6 and 35 of these repeats; affected alleles carry 36 or more. The disease is autosomal dominant, so a single bad copy is enough to cause it, and its symptoms typically begin in adulthood around age 40, after a person may already have had children. Stranger still, the repeat tends to grow longer when it is passed from one generation to the next, which can make the disease appear earlier and more severely in successive generations of the same family. Cystic fibrosis, by contrast, is autosomal recessive, requiring two faulty copies. It arises from mutations in the CFTR gene on chromosome 7, which codes for a channel that moves chloride ions across cell membranes. The most common disease-causing allele worldwide, called delta-F508, is a three-base deletion that removes a single amino acid, the phenylalanine at position 508, from the protein. Because three bases are removed, the reading frame is preserved, but the resulting channel misfolds and never reaches the cell surface, so salt and water balance fails in the lungs and other organs.

The Constant Assault and the Machinery That Fights Back

If mutations could accumulate unchecked, no genome would last a single lifetime intact. DNA is damaged constantly, by ultraviolet light from the sun, by ionizing radiation, by the reactive byproducts of the cell's own metabolism, and by chemical mutagens in the environment. Against this relentless assault the cell maintains a dedicated defense, four major repair systems that patrol the genome and correct most of the damage before the next round of replication can lock it in. There is nucleotide excision repair, base excision repair, mismatch repair, and double-strand break repair, each specialized for a different kind of injury. The fact that you can sit in sunlight, breathe oxygen, and still pass an essentially intact genome to your children is a testament to how well this machinery works. Mutations that cause disease are, in a real sense, the rare failures that slip past an extraordinarily competent quality-control system.

That raises a final puzzle. If most harmful mutations are caught, and the ones that escape make their carriers less healthy and less likely to reproduce, why do disease alleles persist at all? Natural selection should grind them down over generations. For one famous case, the answer is that the very same allele is, in the right circumstances, an advantage. The sickle cell allele reaches frequencies of 10 to 40 percent in regions historically plagued by falciparum malaria, far higher than random mutation could maintain. The reason is heterozygote advantage: people who carry one normal and one sickle allele have substantial resistance to severe malaria, while those with two normal alleles are vulnerable to the parasite and those with two sickle alleles suffer full sickle cell anemia. The carrier, caught in the middle, is the fittest of the three in a malarial environment, so selection preserves the allele rather than eliminating it. This is the classic textbook example of balancing selection, the case where an apparently harmful gene survives precisely because, in a particular world, it also protects.

Key Takeaways

A mutation is any heritable change in the DNA sequence, and despite their fearsome reputation most mutations cause nothing visible at all, because the genome is large, much of it noncoding, and the genetic code is redundant. Mutations sort by scale into point mutations (a single base swapped, with silent, missense, and nonsense outcomes depending on how the change interacts with the code), frameshift mutations (insertions or deletions that are not multiples of three and so scramble every codon downstream), and chromosomal mutations (changes to the structure or number of whole chromosomes, the kind a karyotype can reveal). Specific named alleles produce specific named disorders: sickle cell anemia from a single missense change in beta-globin that swaps valine for glutamic acid, Down syndrome from trisomy 21, Huntington's disease from a CAG repeat expansion in HTT that lengthens across generations, and cystic fibrosis from the three-base delta-F508 deletion in CFTR. Four repair systems catch most DNA damage before it becomes permanent, and where harmful alleles nonetheless persist, balancing selection can explain it: the sickle cell carrier's resistance to malaria is the reason one of medicine's most studied disease alleles has never gone away.