What DNA is

Cell, chromosome, twisted ladder, and reading a real gene

Start the lesson in 3D

Meet a cell

Your body is built from tiny living bricks called cells. There are about 30 trillion of them, far too small to see. Most kinds of cell have a control room inside, called the nucleus. That is where most of your DNA lives. (A few kinds, like the red cells in your blood, have no nucleus.)

An adult human body has roughly 30 to 40 trillion cells. Most types keep a full copy of your nuclear DNA inside a nucleus, a compartment with its own membrane. The orange blobs are mitochondria, the cell’s power plants; they carry a small separate loop of DNA of their own, about 16,500 letters long, inherited from your mother. Mature red blood cells are the famous exception: they lose their nucleus (and their mitochondria) as they mature and carry no DNA at all.

Inside the nucleus

Zoom in. Inside the nucleus is a long, long thread: your DNA. It is the instruction book for building and running you. If you pulled the DNA out of one cell and stretched it straight, it would be about 2 metres long, all squeezed into a space smaller than a speck of dust.

Inside the nucleus the DNA is not loose. It is wound around proteins called histones, forming chromatin, the tangle shown here. A typical human cell nucleus (before it copies its DNA to divide) holds about 6 billion base pairs (3 billion from each parent) and would stretch about 2 metres laid end to end, yet it fits in a nucleus around 6 micrometres across. Most of the time the thread is spread out like this so that the cell can read it.

Chromosomes

DNA does not come in one piece. It comes in 46 pieces called chromosomes: usually 23 from your mum and 23 from your dad. When a cell is about to split in two, it first copies every chromosome, then coils each one up tight. The two copies stay joined in the middle, and that makes this X shape, so they can be shared out without tangling.

Most types of body cell with a nucleus normally have 23 pairs of chromosomes: 22 pairs of autosomes plus the sex chromosomes (XX or XY), one of each pair from each parent. Before a cell copies its DNA, each chromosome is a single, very long DNA molecule; chromosome 1 alone is about 250 million base pairs. The familiar X shape exists only during cell division, after copying: the two identical halves (sister chromatids, one DNA molecule each) are held together at the centromere, the pinch in the middle. The rest of the time chromosomes are unwound and spread through the nucleus.

Unwinding the thread

Pull on the end of a chromosome and it unwinds into a thread. Keep pulling and the thread gets thinner and thinner, until you reach the DNA itself: a twisted ladder.

The packing has several levels. The DNA double helix (2 nanometres wide) wraps around histone proteins to make nucleosomes, like beads on a string about 11 nanometres across. Those fold into thicker fibres and loops, and finally into the compact chromosome, on the order of 10,000 times shorter than the stretched-out DNA. Unwinding it level by level brings you back to the double helix in the next step.

The twisted ladder

Here is DNA up close. It looks like a ladder that has been twisted. The two long sides are the rails. The rungs are made of four kinds of letters: A, T, C and G. A always holds hands with T, and C always holds hands with G. Hover or tap a rung to see its letters. The order of the letters is the message.

DNA (deoxyribonucleic acid) is a double helix. Each rail is a backbone of alternating sugar (deoxyribose) and phosphate; each rung is a pair of bases: adenine with thymine (two hydrogen bonds) or guanine with cytosine (three). Sugar, phosphate and base together make one nucleotide. The two strands run in opposite directions (antiparallel, 5′ to 3′) and are exact complements, which is how DNA is copied: each strand is a template for the other. The model uses idealised B-DNA proportions: about 2 nanometres across, 10.5 base pairs per turn, 0.34 nanometres per base pair, and backbones 144° apart, which is what makes the wide major groove and the narrow minor groove. The tubes are drawn thick so you can see them; it is a cartoon of the shape, not an atomic structure. RNA differs by using ribose, uracil (U) instead of thymine, and usually a single strand. These 32 letters are the start of the protein-coding sequence of your insulin gene.

Words of three letters

To build a protein, the cell reads a copy of the letters three at a time. Three letters make one word, called a codon. Most words mean one building block; a few special words mean STOP. A whole recipe, with all its words, is a gene. Your DNA holds about 20,000 recipes for proteins.

A codon is three consecutive bases. With 4 letters there are 64 possible codons: 61 code for the 20 amino acids and 3 (TAA, TAG, TGA) mean stop, so most amino acids have more than one codon. The protein-coding part of a gene usually begins with the start codon ATG (AUG in the RNA copy, methionine) and is read in frame, three letters at a time, until a stop codon. Many genes also contain non-coding parts: introns, which are cut out of the RNA copy, untranslated regions at each end of the copy, and switches that control when the gene is used. The human genome has about 20,000 protein-coding genes, but their coding parts add up to only 1 to 2 percent of its 3 billion base pairs.

Read a real gene

This is the real start of the recipe part of your insulin gene. Insulin is the hormone that manages the sugar in your blood. Use the arrows to read it one word at a time. Every word becomes one bead: a building block of the protein.

The sequence is the first 90 bases (30 codons) of the coding sequence of the human INS gene, NCBI RefSeq NM_000207. It begins with ATG, so the protein begins with methionine. The first 24 amino acids form a signal peptide that steers the growing chain into the cell’s export machinery and is cut off later; finished insulin is two short chains cut from the rest. Beads are coloured by chemistry: orange hydrophobic, teal polar, blue positively charged, red negatively charged.

Change one letter

Now break it, gently. Tap a letter in the current word and swap it for another. Sometimes nothing changes: the new word means the same block. Sometimes one block changes. Sometimes the new word means STOP, and the recipe ends too early. And if you break the first word, the START, the cell may not begin at all.

A change in the sequence is a mutation. A single-letter change that leaves the amino acid the same is silent (synonymous); one that swaps the amino acid is missense; one that creates a stop codon is nonsense and cuts the protein short. Losing the start codon usually stops the protein being made, though cells can sometimes start from another site. Sickle cell disease is a famous missense mutation: in the beta-globin gene the codon GAG becomes GTG, glutamic acid to valine, at position 6 of the finished protein (word 7 if you count the start codon as this reader does), and the altered haemoglobin makes red cells stiff and sickle-shaped. Adding or removing a letter inside the protein-coding sequence shifts the reading frame, so every word after it is grouped differently; how badly that hurts the protein depends on where it happens.

Copying the recipe

To make insulin, the DNA itself stays inside the nucleus. The cell makes a copy of just that one recipe. The ladder unzips for a moment, a copy of the letters is made (the pink letter is U, which stands in for T), and the copy floats away, out of the nucleus.

This is transcription. RNA polymerase (the pale blob) opens a short stretch of the double helix, reads one strand (the template) and builds a single strand of messenger RNA (mRNA) that matches the other strand letter for letter, with uracil (U) where DNA has thymine (T). The helix closes again behind it. In human cells the new mRNA is then processed (capped, spliced to remove introns, and given a poly-A tail) and exported through the nuclear pores into the cytoplasm.

Building the protein

Outside the nucleus, a tiny machine called a ribosome grabs the copy and reads it word by word. For each word it adds one bead to a growing chain, until a STOP word ends it. When the chain is finished it folds up into a shape. That shape is a protein, and the shape is what lets it do its job. (Only the first ten beads are shown here. The whole insulin recipe makes 110 beads, and the cell later trims that down to the 51 that make up insulin itself.)

This is translation. The ribosome moves along the mRNA one codon at a time. For each codon a transfer RNA (tRNA) carrying the matching amino acid pairs with it, and the ribosome joins that amino acid to the growing chain with a peptide bond, a few per second in human cells. No tRNA matches a stop codon; a release factor protein frees the chain instead. The chain folds, often with help from chaperone proteins, into a precise 3D shape, and the shape decides what the protein does. Shown here: only the first ten codons of insulin, Met-Ala-Leu-Trp-Met-Arg-Leu-Leu-Pro-Leu. The full chain, preproinsulin, is 110 amino acids; processing removes the signal peptide and the middle C-peptide, leaving mature insulin with 51 amino acids in two chains. The folding shown is a cartoon.

Same letters, different proteins

Every recipe uses the same four letters, but in a different order, so the proteins come out different. Haemoglobin carries oxygen in your blood. Collagen makes skin and bones strong. Keratin makes hair and nails. Insulin tells your body what to do with sugar. Actin and myosin pull your muscles. Antibodies fight germs. (The folded shape here is a cartoon, not a real protein.) Tap a chip to see where in the body each one matters.

Proteins are chains of the same 20 amino acids in different orders and lengths, folded into different shapes. Haemoglobin (four chains of about 140 to 146 amino acids) binds oxygen in red blood cells. Collagen, the most abundant protein in the body, forms triple-helix fibres in skin, bone, tendon and cartilage. Keratin builds hair, nails and the outer skin. Insulin, made in the pancreas, is a hormone of 51 amino acids that lets cells take up glucose. Actin and myosin slide past each other to contract muscle. Antibodies (immunoglobulins), made by immune cells, lock on to invaders. Your roughly 20,000 protein-coding genes make far more than 20,000 proteins, because a gene’s RNA can be spliced in different ways and the protein can be modified afterwards. The blob shown here is a schematic fold, not insulin’s real structure.