How a floppy string of amino acids folds itself into a working machine
Every protein leaves the ribosome as a limp chain, yet within milliseconds many snap into an exact three-dimensional shape, and some manage it in a few microseconds. The instructions for that shape, and for the route to reach it, are written into the order of the chain's building blocks, and when folding goes wrong disease can follow.
A freshly made protein starts as a random coil, and folding frequently begins before the ribosome has finished building it, with the front end taking shape while the tail is still being added. The finished form, called the native state, is what makes the molecule useful. Crucially, it depends on the sequence of amino acids rather than merely which ones are present, although surroundings matter too, and similar proteins can fold differently depending on where they sit.
The process runs in stages. First come local patterns such as alpha helices and beta sheets, held by hydrogen bonds, a structure Linus Pauling first described. Anti-parallel sheets are the sturdier kind because their bonds line up at the ideal 180 degree angle. These pieces then pack into a tertiary shape, sometimes locked with disulfide bridges between two cysteines, and in some proteins several folded chains join into a larger quaternary assembly.
The main push comes from water. Oily, hydrophobic side chains force surrounding water molecules into orderly cages, and collapsing those chains into the protein's core frees the water again, raising overall disorder and making folding energetically favourable. Van der Waals attractions in the packed core add stability, while the chain can only bend through a limited set of angles, mapped on a chart called the Ramachandran plot.
Speed varies enormously. Small single-domain proteins of up to about a hundred amino acids often fold in one step, while slow ones outside cells can take minutes or hours, partly because of proline isomerization. Inside cells, chaperone proteins help by steering chains away from wrong shapes and clumps, without being part of the result. Failures can be harmful: misfolded proteins can build amyloid fibrils linked to neurodegenerative illness, infectious forms are known as prions, and many allergies trace back to faulty folding. Simulating all this has challenged computational biologists since the late 1960s.
Source: Protein folding