UCLA

Biological macromolecules such as proteins are remarkable machines at the nano scale. The mechanical motion of proteins that is present in every living system allows those molecules to perform theie specific tasks. Most proteins are enzymes. Enzymes bind to their substrates, speed up the chemical reaction and release the products. Upon this binding and unbinding they undergo a large conformational motion. In case of Guanylate Kinase, the protein under study in my thesis, this deformation is about 1 nm, relatively large compared to the enzyme’s size (~4nm). This conformational motion is essential to the protein’s activity. Proteins are not rigid solids, but rather deformable, wiggling and jiggling inside a medium. This deformability is a general materials property of a folded protein. Enzymes can be deformed and their activity can be modulated by perturbations other than ligand binding.

Fluctuations of enzymes have been studied for a long time and although some progress has been made, still much is unknown about proteins dynamics. In my thesis I try to reveal a little more about proteins dynamics by studying the effect of different kinds of perturbations. For this purpose, a nano-rheology technique is used which allows us to detect deformations as small as half an Angstrom. In this technique we look at the molecule as a whole: the interior polymer chain wrapped by the surface hydration layer. An oscillatory force is applied to the protein, and the amplitude of the resulting oscillation is measured. It has been shown previously, that enzymes are viscoelastic: dissipative at low frequencies and elastic at higher frequencies. I have looked, more closely, for the origin of this atomic scale dissipation.

First, I study the contribution of the hydration layer, which is an integral part of the molecule, by modifying it using an order-inducing compound: Dimethyl Sulfoxide (DMSO), which, at small concentrations is harmless to a protein’s native forms. This compound was added to the molecule’s surface, i.e. the hydration shell, and formed strong hydrogen bonds with water molecules. Even at small DMSO concentrations a significant change in the dynamics was observed: the enzyme became stiffer. The effect is bigger at lower frequencies where enzyme shows dissipative behavior. In that sense, DMSO makes the enzyme more dissipative.

For the first time with this experimental setup, I measured the phase of the response (i.e. the phase difference between the applied force and the resulting deformation, which is basically the imaginary part of the amplitude). Having the ability to measure the phase, we have direct access to dissipation measurements. Phase measurements also revealed that DMSO makes the enzyme more dissipative, which suggest that the hydration layer partially controls the viscoelastic behavior.

Then I investigated at the contribution of the interior, i.e. the polymer chain bulk. To do so I induced point mutations in specific residues of the amino-acid sequence. The mutation points are in a region that goes under a huge strain as the enzyme goes from an open to a closed configuration. Surprisingly, one of the mutated proteins is 10 times more active than the wild type, meaning that the rate of the enzymatic reaction is 10 times faster with the mutated enzyme.

Mechanics of the mutated proteins were studied using the same nano-rheology technique, and some changes were observed, suggesting that both the interior and the surface contribute to dissipation.

I then looked at the deformation caused by ligand binding, in order to see if there is any difference in how different mutants deform upon binding the substrates or unbinding the products. Since the effect of ligand binding is small, our detection method was improved to increase the sensitivity and therefore be able to detect these small deformations. It was seen that some ligands make the enzyme softer while others make the enzyme stiffer. In some cases ligand binding’s signature was only detectable in phase measurements. Overall the conformational change caused by ligand binding was very similar among different mutants, with one exception. For one of the ligands (Guanylate monophosphate-GMP, which makes the enzyme stiffer) the fast mutant seems to have two binding events: one at low concentrations of the ligand and one at higher concentrations, whereas in the wild type only the low-concentration binding event is observed. It seems like the hyperactive mutant might have two binding sites for GMP: at higher concentrations, GMP might bind to ATP binding site. This is a case where specificity is traded off with speed: the mutant is faster but less specific.

Overall, some of my main results are:

- Hydration layer is an integral part of the molecule which is partially responsible for dissipation.

- Both interior and the surface contribute to the viscoelastic behavior. A dissipative contribution in the elastic regime is observed which can be associated with a second dissipation term, namely the one of the bulk.

- Phase measurements provide direct access to dissipation measurements at the atomic scale.