JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

The Double Helix

Two strands twist around a shared axis, their bases paired and stacked inside, their backbones outside — and the shape itself whispers how the molecule could be copied. Here is the most famous structure in biology, and the experiments that revealed it.

From one nucleotide to two strands

You already know the single part: a [[nucleotide-structure|nucleotide]] is a sugar, a phosphate, and one of four bases — A, T, G or C. You also know how these parts string together: the phosphate of one links to the sugar of the next, building a long [[molbio-sugar-phosphate-backbone|sugar-phosphate backbone]] with the bases hanging off the side like beads off a thread. That single chain has a direction, written 5'-to-3'. Now we ask the question this whole rung is building toward: why does DNA come as *two* such chains, wound together? The answer is the double helix — and its shape is not decoration. Almost every property of DNA falls out of it.

Picture a twisted rope ladder. The two long rails are the sugar-phosphate backbones, running up the outside where they face the watery world — sensible, because the phosphates are charged and love water. The rungs reach across the inside, and each rung is a *pair* of bases, one reaching in from each strand to meet in the middle. The whole ladder is given a gentle right-handed twist, so the rails spiral around a common central axis rather than running straight. That twist is what makes it a *helix*; two of them is a *double* helix.

The pairing rule: A with T, G with C

What makes a rung? Not any two bases — only specific partners fit. [[watson-crick-base-pairing|Watson-Crick base pairing]] is the rule that A always pairs with T, and G always pairs with C. The bases hold their partner not with strong covalent bonds but with several weak hydrogen bonds reaching across the gap: A-T is held by two of them, G-C by three. Each pair is a perfect handshake — the right shape, the right hydrogen-bond donors and acceptors lined up. A mispaired rung, say A trying to face G, simply does not fit the geometry and cannot make those bonds.

There is a deeper geometric reason the rule is exactly A-T and G-C. A and G are big two-ring bases (purines); T and C are small one-ring bases (pyrimidines). A rung that works always pairs one big with one small, so every rung spans the same width and the ladder stays an even thickness all the way up. Two purines would bulge too wide; two pyrimidines would fall too short. So the rule is not arbitrary — it is what keeps the helix uniform.

Hydrogen bonds across the rungs are only half the glue. The flat bases are also stacked one on top of the next like a tight pile of coins, and these stacked, water-shy faces cling to each other through what is called [[base-stacking|base stacking]]. Stacking, even more than the hydrogen bonds, is what makes the double helix stable. A handy way to remember the balance: the hydrogen bonds across each rung enforce *which* base pairs with which, while the stacking along the strand supplies most of the *holding power*.

Antiparallel, and grooved

The two strands do not run the same way. They are [[molbio-antiparallel-strands|antiparallel]] — like two lanes of traffic heading opposite ways. Where one strand reads 5'-to-3' going up, its partner reads 5'-to-3' going down. Reading across a rung you always find one strand's 5' end opposite the other's 3' end. This is not a quirk; it is forced by the pairing geometry, because a base pair only fits snugly when the two backbones it joins point in opposite directions. And as you will see, this opposite-direction wiring shapes everything about how the molecule is later copied and read.

    5'- A  T  G  C  A  A  T -3'      <- one strand runs this way
        |  |  |  |  |  |  |          (A-T : 2 H-bonds; G-C : 3 H-bonds)
    3'- T  A  C  G  T  T  A -5'      <- partner runs the opposite way

  read the bottom strand 5'->3' and it spells: A T T G C A T
Antiparallel strands and complementary pairing: knowing one strand tells you the other exactly.

Because the two backbones are not evenly spaced around the axis, the helix's surface is not smooth. The twisting leaves two spiral channels of unequal size running along the outside: a wide [[major-and-minor-grooves|major groove]] and a narrower minor groove. This matters enormously. A protein that needs to recognise a specific DNA sequence usually does not pry the strands apart; instead it reaches a finger into the major groove and feels the unique pattern of chemical bumps that each base pair exposes there. Most protein-DNA recognition — the way a regulator finds the one gene it controls among billions of letters — happens by reading the grooves from the outside, without ever breaking a base pair.

How we found out: the 1953 model

None of this was obvious. In the early 1950s it was not even fully settled that DNA, rather than protein, carried genes. Two clues turned out to be decisive. First, Erwin Chargaff measured the base composition of DNA from many species and found a strange regularity, now called [[molbio-chargaffs-rules|Chargaff's rules]]: in any sample, the amount of A always equalled the amount of T, and the amount of G always equalled C — even though the overall A+T versus G+C ratio differed from species to species. Nobody knew why. With hindsight the answer is obvious: if every A is paired to a T and every G to a C, their totals must match.

Second, Rosalind Franklin, working with Maurice Wilkins, produced extraordinarily sharp [[x-ray-diffraction-of-dna|X-ray diffraction]] images of DNA fibres. Her famous "Photo 51" showed an X-shaped pattern of spots — the unmistakable fingerprint of a helix — and let her measure its dimensions: how far apart the rungs sat, how wide the helix was, how many bases per turn. James Watson and Francis Crick, building cardboard-and-wire models in Cambridge, fit these constraints together. The 1953 model they published was breathtakingly simple: two antiparallel strands, backbones outside, A-T and G-C pairs stacked inside, right-handed twist. It explained Chargaff and matched Franklin's measurements at once.

The shape that explains heredity

The reason the 1953 paper electrified biology was a single understated sentence near its end: the authors noted that the specific pairing they proposed immediately suggested a possible copying mechanism. Here is the idea. Because A only ever pairs with T and G only with C, the two strands are not two independent messages — they are complementary. Each strand carries the full information needed to rebuild the other. The sequence of one strand dictates the sequence of its partner with no ambiguity at all.

  1. Start with the double helix and pull the two strands apart down the middle, breaking the weak hydrogen bonds (the strong backbones stay intact).
  2. Each separated strand now serves as a template — a pattern to read off.
  3. Walk along each template and lay down the only base that fits each one: opposite an A place a T, opposite a G place a C, and so on.
  4. The result is two complete double helices where there was one, each made of one old strand and one freshly built strand — and each identical to the original.

Because each daughter helix keeps one whole old strand and gains one new one, this scheme is called [[molbio-semiconservative-replication|semiconservative replication]] — half of each old molecule is conserved. The structure did not just describe DNA; it *predicted* how cells must copy it, and that prediction was later confirmed experimentally. That is the deepest lesson of the double helix: a molecule's shape can encode not only information but the very recipe for passing that information on. The detailed machinery of copying, reading, and the differences between DNA and RNA are exactly what the rest of this rung and the next ones unpack.