CCR5: structure and dynamics of activation
Two ligands bind the same receptor. They differ by only a few amino acids, yet stabilise distinct activation states.
Ongoing PhD research · Public reference structures (PDB: 5UIW, 7O7F)
01 The target
CCR5: a receptor on the surface of immune cells
Proteins carry out most functions in cells. Some proteins, called receptors, allow cells to sense signals from their environment and translate them into intracellular responses. CCR5 is a transmembrane receptor that guides immune cells by detecting chemical signals during immune responses. It belongs to the GPCR family, the most widely targeted class of drug receptors. Drugs often act by binding to receptors and changing their activity. Because CCR5 regulates immune-cell trafficking, it is a promising target for modifying disease progression in inflammatory and cancer diseases.
In the default camera view, the bottom corresponds to the intracellular side and the top to the extracellular side. The ligand binds at the extracellular face.
02 The chemokines
A few amino-acid changes. Opposite effects.
CCR5 responds to chemokines, small signalling proteins. One example is CCL5, a chemokine composed of about 70 amino acids. Researchers created modified versions of CCL5 by changing only a few amino acids at its tip, the region that inserts into the receptor. Despite these minimal changes, their effects are opposite: one blocks CCR5, while the other activates it more strongly than the natural chemokine.
Tip of the chemokines, first 10 positions (out of ~70 total):
| Molecule | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|
| Blocker | pGlu | Gly | Pro | Pro | Leu | Met | Ala | Leu | Gln | Ser |
| Activator | pGlu | Gly | Pro | Pro | Gly | Asp | Ile | Val | Leu | Ala |
■ Identical ■ Variable positions (5–10)
03 The structure
The same receptor. The same binding site. Very different outcomes.
Both chemokines bind to the same region of CCR5 and insert their tip into the receptor in a similar way. However, the receptor responds differently. When bound to the activating chemokine, one intracellular helix moves outward, allowing signalling proteins to bind inside the cell. With the blocking chemokine, this movement does not occur. This suggests that activation depends not only on where a ligand binds, but on how the receptor changes shape.
Use the toggle above to compare the two structures.
04 Watching it move
From structure to dynamics.
GPCRs are not rigid structures. They fluctuate between many shapes, only some of which can signal. Ligands influence this balance by stabilising particular conformations. The outward motion of the intracellular helix observed above reflects a shift toward an active state.
Molecular dynamics simulations describe how atoms move over time using physics-based calculations. The animation shows CCR5 evolving along a simulated trajectory. The highlighted helix moves outward as the receptor approaches an active geometry. These simulations help us understand not only where a ligand binds, but how it reshapes receptor dynamics and stabilises different functional states.
Methods
Sampling rare transitions
The conformational changes described above often occur on timescales that are inaccessible to standard molecular dynamics simulations. At physiological temperature, a trajectory may remain trapped within a single conformational basin, separated from other states by free-energy barriers. As a result, slow transitions such as GPCR activation may not be observed within practical simulation times.
Enhanced sampling is illustrated here by an adaptive biasing method called metadynamics. This approach constructs a history-dependent bias along a chosen collective variable (CV), a low-dimensional descriptor of the protein’s configuration. By progressively filling previously visited regions of the free-energy landscape, the bias promotes transitions between states that would otherwise be rarely observed.
Because the applied bias is known, free-energy profiles along the selected CV can be reconstructed afterward (illustrative). The underlying principle is closely related to importance sampling: instead of sampling directly from the distribution of interest, we explore a modified distribution that is easier to sample and later recover the unbiased distribution through reweighting. The ΔG slider below sketches how a ligand or mutation shifts the balance between conformational states.
Temperature replica exchange methods follow a different enhanced sampling strategy. Multiple copies of the system are simulated simultaneously at a series of temperatures. Replicas at higher temperatures sample higher-energy configurations more readily, allowing them to cross free-energy barriers and explore a broader range of conformations. At regular intervals, neighbouring replicas attempt to exchange configurations according to a Metropolis criterion that depends on their potential energies and temperatures.
In this way, configurations discovered at high temperature can gradually move down the temperature ladder. Exchanges are accepted probabilistically based on the energies of the exchanging replicas. This allows barriers to be crossed at higher temperatures, while the 300 K replica still samples the same physical behaviour and explores a broader range of configurations than direct simulation would allow.
In my work, I often combine adaptive biasing and replica exchange within one workflow. It makes it quite powerful depending on the system and question.