Nanoscale Dewetting Transition

Studying the possible dewetting transition in protein folding.
Hydrophobicity is believed to be the main driving force in protein folding, a process that still remains a mystery. Understanding the nature of hydrophobic collapse is an important step towards solving the protein folding problem. For simple nanoscale solutes, such as paraffin-like plates, hydrophobicity induces a strong drying transition in the gap between the hydrophobic surfaces as they approach each other. This transition, although occurring on a microscopic scale, is analogous to a first order phase transition from liquid to vapor. The question we try to address here is whether a similar dewetting transition and drying induced collapse occurs when proteins fold or form large multi-protein complexes, and what physical interactions govern the dewetting critical distance as well as the collapse speed.

We first studied the collapse of a two domain protein, the BphC enzyme, which has a large hydrophobic domain-interface area from the hydrophobicity profiling analysis, into a globular structure in order to examine how water molecules mediate hydrophobic collapse of proteins. In the interdomain region liquid water persists with a density 10 to 15% lower than the bulk, even at very small domain separations (such as 4-6 Angstroms). Water depletion and hydrophobic collapse occur on a nanosecond timescale, which is two orders of magnitude slower than found in the collapse of idealized paraffin-like plates. However, when the electrostatic protein-water forces are turned off, a dewetting transition occurs in the interdomain region and the collapse speeds up by more than an order of magnitude. When attractive van der Waals forces are turned off as well, the dewetting in the interdomain region is more profound, and the collapse is even faster. In fact, in this later case, the collapse of the two domains resembles the collapse of two smooth paraffin-like plates both in time scale and in the extent of the dewetting, despite the protein hydrophobic surfaces being much rough. The protein-water electrostatic forces are found to be largely responsible for the much slower collapse in the multi-domain protein than the idealized nanoscale hydrophobic plates, while the van der Waals interactions largely count for the smaller dewetting critical distances (see Figure 1).



The above study shows that it is generally difficult to exhibit such a dewetting transition in real protein systems due to their strong electrostatic interactions with water. To our surprise, we have recently observed such a dramatic dewetting transition inside a nanoscale channel of protein melittin tetramer. Melittin, a 26-residue polypeptide, is a small toxic protein found in honey bee venom, which often self-assembles into a tetramer. The strong dewetting transition occurs in a subnanosecond time scale and a subnanometer (up to 2-3 water diameters) length scale. The dewetting transition is also found to be very sensitive to single mutations of the three very hydrophobic amino acids (isoleucines) to less hydrophobic residues and such mutations in the right locations can switch the channel from being dry to being wet - a "molecular switch". Thus quite subtle changes in hydrophobic surface topology can have a pronounced influence on the drying transition. This study shows that, even in the presence of the polar protein backbone, sufficiently hydrophobic protein surfaces can induce a liquid-vapor transition which can then provide an enormous driving force towards further collapse (see Figure 2).

There appears to be two reasons why the melittin tetramer channel shows a dewetting transition while the previous two-domain protein does not. The melittin channel is like a 1-dimensional tube while the inter-domain region in the two domain protein is a 2-dimensional slab. It is less costly with respect to free energy to disrupt the hydrogen bonds in a tube-like channel. Moreover the unique surface topology springing from the ILE residue sidechains, particularly the ILE2 residues which act as "bumps", destabilizes the wetting in the channel. Even though there is no direct experiment yet showing this nanoscale dewetting transition in real proteins, to our knowledge, there are two very recent experiments showing partial dewetting for paraffin-like systems, one by Steitz et al. for deuterated water (D2O) in contact with polystyrene using neutron reflectivity experiment, and the other by Jensen et al. with water in contact with paraffin using X-ray reflectivity experiment. Partial dewetting with water density about 10% lower than the bulk near the interface was found. It is of considerable interest to devise experiments to test our above predictions for the melittin tetramer.



A better understanding of the dewetting transition might help (1) to design novel water nanopores, such as using carbon nanotubes and superhydrophobic fluorocarbons for molecular water channels (similar to membrane protein Aquaprion); (2) to design nanoscale molecular switches for many important applications, such as drug delivery; and (3) to better understand the mechanism behind all subcellular self-assemblies. As mentioned above, it is of current great interest to design experiments to test our findings from large scale simulations.

Most of these large scale simulations are performed on IBM BlueGene/L development machines. These massively parallel supercomputers such as BlueGene/L provide an opportunity for searching new phenomena in biological universe.


Reference

[1] R. Zhou, X. Huang, C. Margulius, B. J. Berne, Hydrophobic Collapse in Multidomain Protein Folding, Science 305, 1605, 2004

[2] P. Liu, X. Huang, R. Zhou and B.J. Berne, Observation of a Dewetting Transition in the Collapse of the Melittin Tetramer, Nature 437, 159-162, 2005

Contact: Ruhong Zhou <ruhongz@us.ibm.com>



Last updated 7 Jul 2006