Running with KITA

A new clue about the folding of proteins comes from studies with a novel technique known as kinetic terahertz absorption spectroscopy (KITA). Results have now been compared for the first time with X-ray diffraction (SAXS) results, and fluorescence and circular dichroism spectra to reveal how protein folding takes place in two stages on different timescales.

Protein folding is one of the great conundrums of the twenty-first century. How exactly does a linear string of amino acids "know" into what three-dimensional cross-linked structure to fold itself? Moreover, how might molecular biologists predict this folding from first principles and how might the misfolding seen in prionic diseases, Alzheimer's and other disorders be prevented or even reversed?

Martina Havenith of Ruhr-University Bochum, Germany, and Martin Gruebele of the University of Illinois Urbana-Champaign, and their respective research groups have together produced a picture of the protein folding process as it occurs in aqueous solution using KITA. In their studies, they used the ubiquitous protein, known not by coincidence as ubiquitin. This is a 76-residue, predominantly beta-sheet protein that has long been used as a prototype for folding kinetics studies. The researchers were able to record the folding process with a temporal resolution of one image per millisecond.

"Recently, there has been a growing interest in probing not just the dynamics of self-assembling macromolecules but the dynamics of their solvation shells as well," explain the researchers. They point out that dielectric, Raman, fluorescence, and NMR spectroscopic techniques as well as neutron scattering, and crystallography have all been applied to providing new insights. However, spectroscopy at terahertz frequencies (equivalent to a wavelength range of 0.1-1 mm; 1 THz equals 1 ps^-1) could directly probe picosecond solvent dynamics on any timescale. It is, the researchers add, surprising how sensitive a protein is to hydration layers far from the molecular surface, as opposed to those water molecules only in its immediate vicinity, something that was revealed by the KITA studies. By comparing the KITA data with SAXS, fluorescence and CD spectra they could also see that the protein folding took place in two phases.

The team explains that in a very rapid first phase, the protein collapses in less than a millisecond, while simultaneously, a rearrangement of the protein-water network takes place. In a slower second phase, complete after almost one second, the protein folds to its native state. It is only now, with KITA, that THz spectroscopy has gained access to anything but steady-state observations. "Only now can we see the whole stage play," explains Havenith, "no longer just the opening scene and the curtain call.

Previously, the RUB and Illinois team had demonstrated how much of an influence a protein could have on its aqueous environment. In the bulk, hydrogen bonds between individual water molecules form and break every 1.3 picoseconds but even a small amount of protein will increase order in the liquid, altering the dynamics of the solution. Whether or not the protein in question is folded or not has a different effect on the ordering of water molecules, something that KITA spectroscopy can lay bare even in such a short time period.

By starting with an unfolded protein in aqueous solution and then initiating the folding process, the chemists could monitor changes as they occurred using KITA spectroscopy. Apparently, within ten milliseconds, the motions of the water network are altered as well restructuring of the protein itself occurring. "These two processes practically take place simultaneously," Havenith adds, "they are strongly correlated." After these initial changes, a second significantly slower phase takes place within the protein itself. In this process, the protein folds to its final native conformation.

"Conceptually, the interpretation of the THz kinetics experiment is straightforward," the researchers say, "For the 2-3 THz data, we proposed a coupling of protein surface flexibility and hydration shell to explain the sensitivity of absorbance to side-chain truncations in the core of the protein. In contrast, KITA kinetic measurements at 0.1-1 THz are not sensitive to changes in protein flexibility that result from side-chain truncations, because the protein is just beginning to fold and does not yet have a well-defined surface." This, they say, suggests that a different mechanism influences the THz spectrum at lower frequencies early during the folding process.

The results provide new evidence for the simple, yet controversial, idea that water molecules play a wider more fundamental role in protein folding, and thus in protein function, rather than being a passive back drop against which protein dynamics are played out.