Proteins can perform a remarkable variety of functions, including the catalysis of biochemical reactions, cell signalling and the transport of molecules. However, Hans Frauenfelder and colleagues suggest that, on their own, these 'picomachines' are essentially stalled — what brings them to life is the surrounding water, which acts as a kind of puppeteer controlling the protein dynamics (Proc. Natl Acad. Sci. USA doi: 10.1073/pnas.0900336106; 2009).

Credit: © SPL

The interaction between a protein and its environment happens on two levels: on the one hand there is the bulk solvent, on the other the hydration shell (made up of the water molecules that interact directly with the protein). In their 'unified model of protein dynamics', Frauenfelder et al. propose that fluctuations in the hydration shell dictate internal protein motions, whereas fluctuations in the bulk environment control large-scale motions of the protein, such as shape changes.

In water, as in other glass-forming liquids, there are two qualitatively different types of equilibrium fluctuation, which originate from the bulk solvent and the hydration shell, respectively. Furthermore, the energy landscape of proteins — which defines the conformations a protein can assume — is organized hierarchically. To take account of both of these factors, Frauenfelder et al. associate the two types of fluctuation with random walks in different tiers in the highly structured energy landscape. As the fluctuations in the bulk water are controlled mainly through its viscosity (which is believed to be quite constant in cells), the range of possible large-scale motions that these fluctuations can control seems relatively narrow. The hydration shell, in contrast, is highly structured, implying that many possible internal motions can be induced.

Frauenfelder et al. have compared their predictions with experimental data on the dynamics of the biomolecule myoglobin (pictured), and have found them to be consistent. Their model correctly predicts the motion of carbon monoxide through this protein and the relaxation of the protein structure after the carbon monoxide has been released. Nevertheless, they advise caution because there are other motions, not only the dynamical processes considered in their study, that are important in protein function.