Skip to content
Spark & Whetstonethe grown-up commons

← The Shape of Life

The Shape of Life

Four short lessons on proteins, the folded chains that do most of the work in living things: why shape matters, how a chain folds, what a prediction can and cannot tell you, and how pictures of the real thing are made. Pairs with the Fold Commons Viewer.

Spark & Whetstone · sparkwhetstone.org/learn/shape-of-life · CC BY-NC-SA 4.0. Answers are at the end, so you can try each question first.

Lesson 1. Shape is the job

A protein is a chain of smaller building blocks, folded into a particular shape. The shape decides what it does: a pocket that holds one partner, a channel that lets one thing through, a pair of arms that grips.

Change one building block in the wrong place and the shape can change, and with it the job. Sickle-cell disease comes from a single changed block in haemoglobin.

That is why so much of biology is about knowing shapes.

Question. Why can a single changed building block matter so much?

  1. Because it makes the protein heavier.
  2. Because it can change the shape, and the shape decides the job.
  3. Because proteins only have one building block.

Say it back. In your own words: why do biologists care so much about the shape of a protein?

________________________________________________

________________________________________________

Lesson 2. How a chain folds

In water, the oily building blocks of a chain end up packed together on the inside, away from the water. This alone pulls the chain into a compact shape.

Along the way, parts of the chain settle into common local shapes: spirals called helices, and flat strands that lie side by side as sheets.

Proteins do not decide to fold. The pushes and pulls of water and of the chain itself settle it, usually in a fraction of a second.

Question. What mainly pulls oily building blocks to the inside of a folded protein?

  1. The protein wanting to protect them.
  2. The surrounding water, which holds on to itself and crowds oily parts together.
  3. Gravity.

Say it back. Say it back: what is the hydrophobic effect, in a sentence a child could follow?

________________________________________________

________________________________________________

Lesson 3. A prediction, and how sure it is

Since 2020, computer programs such as AlphaFold have predicted the shapes of hundreds of millions of proteins from their sequences alone.

A good prediction says how confident it is about each part. On the Fold Commons Viewer, confident parts look firm and less confident parts look faint, like a pencil sketch.

Low confidence is not always a mistake. It often marks a floppy part that really does move, and some of those floppy parts do important work.

Question. A predicted structure shows one long tail with very low confidence. What is a reasonable reading?

  1. The prediction is broken and should be ignored.
  2. The tail may be floppy in real life; low confidence can mark a part that moves.
  3. The tail is certainly a helix.

Say it back. In your own words: what does a faint, low-confidence region in a prediction tell you?

________________________________________________

________________________________________________

Lesson 4. Pictures of the real thing

Experiments show what is really there. One common method freezes many copies of a protein and images them with an electron microscope, then combines thousands of pictures into one shape.

Where the copies all look alike, the picture is sharp. Where they differ, because a part moves, the picture is blurred. Experiments can also show partners, such as a small molecule in a pocket, that a prediction from one chain cannot.

Prediction and experiment are used together: each is honest about a different weakness.

Question. Why might an experiment show something in a pocket that a prediction did not?

  1. Because predictions are always wrong.
  2. Because the partner is not part of the protein's own chain, so a prediction from the chain alone cannot draw it.
  3. Because microscopes add things.

Say it back. Say it back: what can an experiment show that a prediction cannot, and the other way round?

________________________________________________

________________________________________________

At the table: Sketch, then photograph

Players: 2 or more, any ages from about 9. You'll need: A pen, a pencil, paper and a phone camera.

  1. Each player draws a familiar object from memory, such as a bicycle or a kettle: pen for the parts you are sure of, pencil for the rest.
  2. Photograph the real object.
  3. Lay drawing and photo side by side and mark what each got right and what each missed.
  4. Then open one protein in the Fold Commons Viewer and find its firm parts and its faint parts.

Seat sheets and a run-through: sparkwhetstone.orghttps://foldcommons.org/

Answers

  1. Because it can change the shape, and the shape decides the job. — A protein's job follows from its shape, so a change that alters the shape can alter the job.
  2. The surrounding water, which holds on to itself and crowds oily parts together. — This is the hydrophobic effect: water molecules stick to each other, so oily parts end up gathered together, inside the fold.
  3. The tail may be floppy in real life; low confidence can mark a part that moves. — Very low confidence often coincides with regions that are disordered in reality, so it is information worth reading, not just noise.
  4. Because the partner is not part of the protein's own chain, so a prediction from the chain alone cannot draw it. — A prediction from one chain shows that chain. An experiment images the real sample, partners and all.