Protein folding is a fascinating topic for studying imagination in science because scientists have created a common visual field for proposing and testing their theories. They call this field an “energy landscape,” and the shape of the landscape — its hills and valleys — reflect how scientists conceptualize the ways that proteins acquire their functional, productive form in our bodies.
Energy landscapes can be represented mathematically and visually, but I will focus on the latter here. Each protein has its own energy landscape, which can be represented visually as a diagram using scientific data. But the same kind of diagram can be used generically, too, to represent common mechanisms for how proteins fold across every protein.
You may have noticed the metaphor of proteins “folding.” My grandma always thinks of it like origami. This is not a bad place to start, except that proteins are more like long, thin strips of paper than big flat sheets. Folding a protein, then, involves bending the strip along many angles and in many places to collapse the protein into a dense cluster. Here’s where the analogy starts to break down, though, because an origami crane, for example, keeps its shape because we irreversibly crease the paper. Proteins are made up of small molecules called amino acids that are bonded together like beads on a string, and while the bonds between amino acids can bend they don’t crease. What keeps a protein in its folded shape, then, are attractions and repulsions between amino acids and the water around them.
To improve on the origami analogy, we might think of an unfolded protein as a string of magnets laid out in a strong magnetic field. Do you remember how iron filings line up in different patterns if you pour them out onto a glass table and move a magnet around underneath? The iron filings aren’t magnets, so they just rotate in a way that lines up with the strength of the magnet under the table. (The filings become parallel with the magnet’s field lines.) Now imagine that we do a similar experiment but with a bunch of little magnets on a string instead of loose iron filings. Magnets have two poles, north and south, and opposite ends attract while similar ends repel each other. If we put a big magnet under a table with a bunch of little magnets on a string, and the north pole of the big magnet is facing up, then the little magnets have a problem. Each of them have a north and a south pole, but only the south pole is attracted to the big magnet under the table while the north pole is repulsed. Because all the magnets are stuck together on a string, they won’t be able to move individually to minimize their conflict with the big magnet.
In addition to the overall effect of the big magnet, there are also interactions between each of the small magnets. Each pair would be better off aligning their north and south poles, but again because they are stuck on a string together it’s not possible to optimize every pair simultaneously. One pretty good solution would be to “zip” up the magnets as pairs along the string, sticking together the magnets at opposite ends of the string and then connecting each corresponding pair inwards.
This analogy is more precise than the origami idea because it captures how the environment of the protein affects its shape. Instead of a big magnet, a real protein is surrounding by many hundreds or thousands of water molecules (and in a real cell, many other proteins, sugars, ions, and other stuff). While magnets all have a north and south orientation, the amino acids come in 20 different types. Each pair of types might be either attracted or repulsed from each other (instead of potentially both) or neutral. This makes “zipping” up the protein much trickier, because the different types of amino acids can be positioned anywhere along the string and there’s no guarantee we’ll have the right number of attracted pairs.
We’re now in a position to understand the meaning and the complexity of the energy landscape idea. Imagine holding a protein in your hands, with a random set of amino acids from the 20 types arrayed along it like beads. Every pair of amino acids along the string can be attracted, repulsed, or neutral toward each other. Imagine that for any configuration you put the string in, we can calculate the net amount of attraction and repulsion between all the pairs (we combine the magnitudes of the interactions for every pair). Now add in the fact that there are similar attractions and repulsions between the water molecules surrounding the protein and each amino acid.
Putting these all together, we get a single number which summarizes the energy level of the protein — is this configuration very stable and well-balanced, so that repulsions cancel out attractions? Or is it very unstable, so that attractions outweigh repulsions or vice versa? Stability means a low energy, while instability means a high energy. This energy level is the height of the landscape for each configuration of the protein.
Remember the kind of bumpy maps where mountains actually push up in 3D out of the paper? (They’re called raised relief 3D topographic maps.) When there are a bunch of configurations of the protein that are very similar and have a high energy level, we get a mountain in our energy landscape. Similarly, when a bunch of shapes cluster together with low energy, we get a valley. You could imagine tracing a path (going for a hike) along the energy landscape by moving the string around into different positions, travelling over mountains when the protein is unstable and strolling through valleys when it’s stable.
Protein folding, then, is the process the protein goes through naturally (without us guiding it) as it moves from some position high up in the energy landscape down into one or another valley. It’s not as if the protein can see where it’s going, of course, like we can looking at the whole landscape as a map. Yet proteins manage on the whole to find the lowest positions in their energy landscape incredibly quickly, usually on the order of microseconds (millionths of a second). So understanding how proteins fold in nature is a fascinating problem for scientists — is the protein’s trick having a special path through the landscape that it always follows, or always starting with a smooth landscape that always tilts downward toward the most stable point? In the next post I’ll collect some images of energy landscapes from scientific papers.