Thursday, February 7, 2008

Some old, lots new

I seem somehow to have overlooked a couple of highly relevant review articles in my Chem. Rev. paper. One is Ken Dill’s piece on ‘Modeling water, the hydrophobic effect, and ion solvation’:
K. A. Dill, T. M. Truskett, V. Vlachy & B. Hribar-Lee, Annu. Rev. Biophys. Biomol. Struct. 34, 173-199 (2005).
(Get it here.)
The other is Martin Chaplin’s paper ‘Do we underestimate the importance of water in cell biology?’, which, as the title implies, provides a much briefer overview of most of the issues I discuss in my review and carries the same basic message:
M. F. Chaplin, Nature Rev. Mol. Cell Biol. 7, 861-866 (2006).
(Get it here.)
Apologies for these omissions.

Moving on to things new… David Chandler’s comments on the mechanism of dewetting-induced hydrophobic assembly in my last post are expanded on in some detail in a new paper (A. P. Willard & D. Chandler, J. Phys. Chem. B doi:10.1021/jp077186+). They argue that the motions of the hydrophobic solutes are such that the basic cavitation process to form the ‘vapour bridge’ can have zero activation energy.

One of my favourite papers of the moment comes from Martina Havenith at Bochum and colleagues (S. Ebbinghaus et al., PNAS 104, 20749-20752; 2007). They have used terahertz spectroscopy and MD simulations to probe the hydration layer around proteins, in particular to estimate how thick it is. They find that differences in correlated water motions, relative to the bulk, extend to more than 2 nm from the protein surface. It’s a remarkable and important demonstration of the extra ‘reach’ that solvation affords biological macromolecules, and really makes the case for why the hydration layer needs to be considered in some sense a part of the molecule it encompasses, making them fuzzy-edged entities with a sphere of influence that stretches well beyond the apparent surface.

Martina and her colleagues have also looked at how this solvation structure is altered by mutations and by pH (S. Ebbinghaus et al., JACS doi:10.1021/ja0746520; paper here). They find that a single mutation of the five-helix bundle lambda*[6-85], replacing a glutamine side chain with aromatic residues, significantly reduces the reach of the perturbation to the solvation water. This distance is also reduced when the wild-type protein is denatured at pH 2. It will be interesting to know if both are general effects, implying that proteins are somehow optimized to induce maximum restructuring of the solvent.

Bertil Halle has sent me a preprint of a paper just accepted in PNAS, in which he along with Johan Qvist and Monika Davidovic at Lund, and Donald Hamelberg at UCSD, report a wholly water-free large hydrophobic cavity in bovine beta-lactoglobulin. This cavity, called the calyx and acting as a binding site for fatty acids and other nonpolar ligands, has a volume of 315 cubic Å. The authors contrast this with the interior of carbon nanotubes, which is nominally hydrophobic but threaded by water chains. It shows just how amazingly dry nature can keep itself when the need arises.

Bertil also has a new paper) with Erik Persson describing how magnetic relaxation dispersion studies of water molecules buried inside proteins can provide a probe of ns- to ms-timescale protein dynamics (E. Persson & B. Halle, JACS 130, 1774-1787; 2008).

Talking of water in nanotubes, Hideki Tanaka and his colleagues at Okayama University have ised MD simulations to map out the complete phase diagram of water in nanotubes at atmospheric pressure for diameters up to 1.7 nm (D. Takaiwa et al., PNAS 105, 39-43; 2008; paper here). They find at least nine different ice phases, each apparently adapting to the confined space in a way that maximizes the number of hydrogen bonds. They say that the confined liquid water doesn’t show a density maximum above freezing point, and that it shrinks on freezing. Nor is freezing necessarily a first-order transition here. All a rather beautiful picture of how profoundly confinement can alter water’s properties.

There’s a curious paper in Langmuir (A. P. Sommer, A. Caron & H.-J. Fecht, Langmuir 24, 635-636; 2008) claiming that ‘ordered interfacial water’ near hydrophobic and hydrophilic surfaces can be tuned with laser light, and that the light causes an increase in fluidity, presumed (as far as I can make out) to be due to depletion of the ‘ordered’ layer, in the hydrophilic case. I’m left wondering whether there is really any direct evidence for ‘increased ordering’ in these interfacial layers, and what precisely that means here. Something interesting seems to be happening, but I don’t think it’s clear what it is.

Aizhuo Liu et al. in Michigan report what they call ‘bifurcated’ hydrogen bonds in proteins, using isotope substitution studied with NMR (A. Liu et al., JACS doi:10.1021/ja710114r). But as far as I can see, these are not the bifurcated hydrogen bonds postulated by Sciortino et al. to play a role in molecular mobility in the liquid state (Nature 354, 218; 1991), where one proton binds to two oxygens. Rather, what we have here are simply oxygen atoms linked to two protons via H-bonding. Can we clean up the terminology please?

A paper by Masahide Terazima at Kyoto and colleagues offers evidence for the role of hydrophobic interactions in light switching of the antirepressor AppA of Rhodobacter sphaeroides (P. Hazra et al., J. Phys. Chem. B 112, 1494-1501; 2008. This molecule forms a dimer when its photosensitive BLUF domain is activated by blue light, and this photoactivated state then represses expression of genes involved in photosynthesis. It’s a rather subtle example of hydration changes inducing a biological behaviour.