The Hydrophobic Effect
Predicting a protein’s structure and function from only its sequence is one of one of the most significant problems in modern computational biology. This “protein folding problem”, as it is known, has been in the news recently with DeepMind’s groundbreaking AlphaFold2 results at CASP14. This progress comes after decades of research and attempts at the problem by hundreds of research groups. A key part of what makes this problem so difficult is that proteins don’t fold by themselves, but by complex interactions with the surrounding environment, water. While we might take water for granted as the background where other chemistry takes place, it actually plays an active role in protein folding. Water is vital as the driver of hydrophobic packing, in which the core of a protein forms. This is an example of the hydrophobic effect, a property of how hydrophobic or “water-fearing” molecules interact with water. This effect is one of the keys to protein folding, but it can’t be modeled directly, which makes it extremely difficult for scientists to predict a protein’s structure.
In a previous post, we discussed what proteins are and the roles they play in biology. Proteins are made up of long chains of amino acids linked together. Think of a long string of beads, with each bead being a different amino acid. Although this makes it sound like proteins are relatively boring and linear, we know that this isn’t the case. In the real world, proteins take compact and interesting shapes that define their roles in a biological setting. Going back to the analogy of beads on a string, this means that a string of amino acid beads won’t stay linear — it will spontaneously clump itself together.
The hydrophobic effect is the main thing that drives proteins to fold over themselves into the right shape. Although it may sound unfamiliar, we actually encounter the hydrophobic effect every day. The hydrophobic effect is why oil and water don’t mix, why it’s difficult to wash off oil without soap, and why putting out grease fires with water is a very bad idea. It can also be used to our benefit. For example, there are hydrophobic sprays that can be applied to glasses and other surfaces to prevent them from fogging or getting wet. As strange as the hydrophobic effect may seem, it can be understood through the lens of entropy — a physical property that describes uncertainty. More intuitively and less accurately, for a given system, which can be anything from a single atom to an entire universe, entropy tells us how disordered it is. Along with entropy is the second law of thermodynamics; the universal rule that every system eventually reaches its maximum entropy and falls into randomness and disorder.
When molecules are surrounded by water, they will tend to interact with it if they can. For polar molecules, those with partially positive or negative charges, this isn’t a problem. This class of molecules will typically dissolve and have momentary ionic interactions with water. Sugar is an example of a polar molecule that does this. However, nonpolar molecules like fats and oils can’t interact with water this way. Instead of freely interacting with the water, these nonpolar, hydrophobic fat molecules actually dampen the movement of the water molecules. Surprisingly, when confronted with a hydrophobic molecule, water molecules tend to gather around it in a highly-ordered arrangement known as a “clathrate cage”.
Water molecules in a free solution move chaotically, bouncing in random directions determined by countless random collisions with other water molecules. At an interface with a hydrophobic molecule, these random collisions are biased in one direction — towards the hydrophobic molecule — which keeps water molecules at the interface instead of returning to the solution where they can continue their chaotic bouncing. Additionally, water molecules stuck at this interface will tend to interact with each other to form hydrogen bonding networks around the surface of the hydrophobic molecule — creating the clathrate cage.
Water molecules forming clathrate cages tend to have their movement severely constrained, as they can’t easily return to the bulk solution. Water molecules in solution can move in all three directions in space, but the ones on the interface of a hydrophobic molecule are effectively limited to moving along the two-dimensional surface of the interface. Being forced into a highly ordered structure that removes their random motion means that water molecules in a clathrate cage have less entropy than the ones in the rest of the solution, because their arrangement is now less uncertain. Because larger hydrophobic areas need larger clathrate cages, they end up being more ordered. Accordingly, small hydrophobic interfaces have smaller clathrate cages. Smaller cages mean there is less order, more chaos, and the system is entropically favorable. What we see, then, is that hydrophobic molecules will stick together in an “effort” to minimize the area between themselves and water. This leads to fewer water molecules in an ordered state, and satisfies the universal rule that the system should be as disordered as possible!
So why do systems have this tendency to maximize entropy? Entropy roughly measures how much is unknown about a given system. Specifically, it tells us roughly how many possible configurations of the system are available to us. For example, a gas like air will fill its container in an effort to maximize entropy — the more possible locations for its molecules to be in, the higher the entropy of the gas. Another example is a pair of headphones getting tangled in your pocket. There are many possible ways for the headphones to be tangled up and relatively few ways for them to not be, so they will naturally become tangled.
For proteins in solution, hydrophobic residues such as valine or phenylalanine will be susceptible to the hydrophobic effect and will tend to aggregate together into the hydrophobic “core” of the protein. Many mutations which cause proteins to misfold are mutations that destabilize their hydrophobic cores, or lead to proteins hydrophobically sticking together when they shouldn’t be. A famous example of the last situation is sickle cell anemia, in which a mutation changes just one surface amino acid of hemoglobin. Instead of glutamate — a charged residue which interacts well with water — some people have a hydrophobic valine amino acid on their hemoglobin. This small change causes hemoglobin proteins to aggregate together into long filaments, pushing and stretching cells from the inside, and forming the sickle-shaped red blood cells associated with the disease.
Understanding the hydrophobic effect is a significant hurdle for scientists attempting to model protein folding computationally. It is unique because it doesn’t come directly from the known forces of nature, but instead arises indirectly from the complicated interactions between proteins and water. As our understanding of the hydrophobic effect improves, so too will our understanding of proteins and their structures.
Citations:
Compiani, Capriotti (2013). Computational and Theoretical Methods for Protein Folding. https://doi.org/10.1021/bi4001529