Neutrons can observe proteins at work

Proteins are biological macromolecules and are found in all cells. There they provide structural strength or, in their role as molecular machines, enable metabolic processes to take place, such as the transportation or utilization of energy sources. The function of certain proteins depends on the accurate recognition of substrates and the reshaping of the protein structure to make the final product. A range of neutron scattering methods offers unique opportunities to understand the structure and movement of protein components in action.

Proteins are long-chain macromolecules. They are made up of chains of amino acids in a particular sequence and have a characteristic three-dimensional arrangement Together, these determine their specific chemical and physical properties. A key molecular feature of many proteins is the ‘active site’, in which a catalytic reaction takes place. This is a region where a perfectly matching small molecule called a “substrate” latches onto the protein to induce a structural or chemical change. The structural changes range from the reorientation of atoms through to the rearrangement of complete domains. These can enclose substrates, release products or reconfigure domains.

Neutron scattering provides a unique non-destructive tool for probing the behaviour of proteins in their nearly-natural aqueous environment. For instance, small angle neutron scattering, (SANS), reveals the arrangement or overall shape of domains, and neutron spin echo (NSE) spectroscopy enables the observation of dynamics over the relevant length and time-scales – nanometres and nanoseconds – during the reorientation of complete domains. Neutron crystallography applied to protein crystals using the Biodiff instrument makes it possible to locate hydrogen atoms at a resolution of up to 0.2 nanometres, as opposed to X-ray crystallography, which is sensitive to electron density distributions and cannot detect hydrogens. Backscattering-techniques are capable of identifying atomic motions on the scale of interatomic distances and picosecond timescales.

Using phosphoglycerate kinase (PGK [1], see Figure 1) as an example, it is clear how different neutron scattering methods complement each other. PGK catalyses a step in the glycolytic pathway and is used to win energy for the cells from sugar. PGK has a widely open domain structure with a hinge near the active centre between the two domains. Two key amino acids, important for phosphoryl transfer, are located in both the red and blue domains.

Structural analysis using SANS (Fig. 2) revealed that the structure in solution is more compact compared to the crystal structure, but does not allow the functionally important phosphoryl transfer between the substrates, as the separation between the two amino acids is too great. Experiments using NSE showed thermally-driven large-scale domain fluctuations on a timescale of 50 nanoseconds. Could these intrinsic fluctuations of the protein be strong enough to cast the protein into an active configuration, whereby the two key amino acids in the domains come close enough together to initiate reactions involving the bound substrates?

A comparison of the dynamics with the slowest displacement patterns in a normal mode confirmed this. A normal mode analysis is a calculation of all possible movements of a protein, based on atomic structure information. The slowest motional patterns which also have the greatest amplitudes, exhibited dynamics that fit perfectly with the dynamics observed. As the slowest normal mode facilitates a close encounter of the key residues in the active centre to bring about the active configuration, the observed dynamics in fact enables the functional capability of the protein.