Surviving the winter

How do some organisms survive dehydration? In the state called anhydrobiosis, such organisms exhibit a kind of suspended animation, showing no apparent metabolic activity but reviving when rehydrated. Two explanations for this behaviour have been proposed: replacement of water by another medium, such as sugars, and the formation of a glassy matrix (vitrification). Takashi Okuda of the National Institute of Agrobiological Sciences in Tsukuba, Japan, and his colleagues have reported evidence that, at least for the case of the African midge Polypedilum vanderplanki, both these hypotheses seem to apply [M. Sakurai et al., PNAS 105, 5093-5098; 2008 – paper here]. They find that anhydrobiotic larvae of this insect are in a glassy state in which much of the water is replaced with trehalose. They think that the glassy matrix is a mixture of trehalose with other components, such as highly hydrophilic proteins. The big question is surely how the organisms get into and out of the glassy state: if moisture uptake or high temperature turns the sugar glass rubbery, the larvae lose their viability.

The production of trehalose and other compatible solutes is of course one of the common mechanisms of freeze-tolerance. Cold conditions threaten organisms not just because of the physical disruption caused by ice crystal formation, but because proteins may denature below about minus 20 Centigrade. The mechanism of cold denaturation has been much debated. Cristiano Dias of the University of Montreal and colleagues propose [C. L. Dias et al., Phys. Rev. Lett. 100, 118101; 2008 – paper here] that it shares characteristics with pressure-induced denaturation, which Gerhard Hummer and his coworkers have previously attributed to the destabilization of hydrophobic contacts in favour of solvent-separated ones [G. Hummer et al., PNAS 95, 1552; 1998]. In other words, the idea is that both effects are all about the temperature- and pressure-dependence of the hydrophobic interaction. Dias and colleagues support this conclusion with 2D molecular-dynamics simulations, which suggest that at low temperatures water molecules in a protein’s hydration shell hydrogen-bond more strongly than those in the bulk.

It’s a bit of a month for studies of freeze/dehydration tolerance. Martina Havenith at Bochum and colleagues have used terahertz spectroscopy to show surprisingly long-ranged correlations between sub-picosecond dynamics in water close to and further from dissolved cabohydrates (trehalose, lactose, glucose) [M. Heyden et al., JACS doi:10.1021/ja0781083– paper here]. That is, the effects persist over about 5-7 Å, suggesting that this is in effect the width of the hydration shell for these sugars. The perturbations are stronger for the disaccharides than the monosaccharide (here the influence extends only to about 3-4 Å), which the authors connect with the stronger cryoprotection activity of the former – that is, it would support the notion that this protection is conferred by a disruption (retardation) of the solvent dynamics.

And Giovanni Strambini and coworkers at the Consiglio Nazionale delle Ricerche in Pisa, Italy, have used fluorescence spectroscopy to study the stabilization of the azurin protein by sugars and polyols in ice [G. B. Strambini et al., J. Phys. Chem. B 112, 4372-4380; 2008 – paper here]. One question they have set out to address is: to what extent are the cryoprotectants defending against low temperature per se, and to what extent against ice formation? In fact, none of the molecules (sucrose, trehalose, sorbitol, glycerol) seems to fully protect against ice formation, although trehalose ‘tries hardest’. It seems that the role of the cryoprotectants is not to increase the thermodynamic stability of the native fold per se, but to help hold it together specifically in the liquid phase. Curiously, some of them do this less effectively at minus 15 C than at lower temperatures – which seems to be because once it gets cold enough, denaturation becomes kinetically rather than thermodynamically controlled: unfolding is simply too slow. There is an interesting story building here.

More on low-temperature protein dynamics comes from Jeremy Smith and his coworkers Vandana Kurkal-Siebert and Ritesh Agarwal at Heidelberg [Phys. Rev. Lett. 100, 138102; 2008 – paper here]. They have used MD simulations to look at interprotein dynamical interactions in hydrated crystals of carboxymyoglobin, a kind of proxy for protein-protein interactions more generally. In fully hydrated crystals there is a dynamical transition at around 240 K due to intermolecular fluctuations, whereas no such behaviour is seen for low hydration. This is reminiscent of the 220 K intramolecular dynamical transition seen in proteins, which resembles (but may not in fact be – see below) a glass transition. If I’ve understood this properly, the idea is that the (diffusive) intermolecular motions in the proteins must therefore be mediated (activated) by those in the hydration shells.

The dynamics of proteins in this pseudo-glassy state below about 220 K do indeed seem to get very closely coupled – perhaps ‘slaved’ – to the similar non-Arrhenius dynamics of the solvent. Now Martin Weik and a host of others have confirmed this coupling using neutron scattering [K. Wood et al., JACS 10.1021/ja710526r – paper here]. They find that below the transition temperature, the protein (here maltose binding protein) is structurally arrested in a glassy solvent cage, while at around 220 K ‘fast’ protein motions are reawakened at precisely the same time as hydration water starts to show diffusive translational motion. In a nutshell, as the authors put it, “the protein dynamical transition is correlated with relaxation of the protein H-bond network, which, in turn, is associated with the onset of water translational diffusion.”

Sow-Hsin Chen at MIT and his coworkers have previously related this dynamical transition to that in pure water predicted on the basis of a liquid-liquid transition at high pressure, which creates a kind of ‘ghost’ of the transition called the Widom line at lower pressures (see S.-H. Chen et al., PNAS 103, 9012; 2006). They now report further evidence, from SANS studies of water in mesoporous silica (MCM-41), of a change in the structure and dynamics of water at around 235 K due to crossing of the Widom line [D. Liu et al., J. Phys. Chem. B 112, 4309-4312; 2008 – paper here] .

Finally (for now), Roland Netz and colleagues at the Technical University of Munich have attempted to simplify studies of the hydrophobic attraction, which has typically involved mesoscopic surfaces on which many length scales can be important, by investigating the force required to peel a single, mildly hydrophobic peptide from a diamond surface [D. Horinek et al., PNAS 105, 2842-2847; 2008 – paper here]. They compare their experimental results using atomic force spectroscopy with simulation, to try to quantify the relative contributions of dispersion forces between the two surfaces and ‘water structure’ effects. They find that all the individual pairwise interactions between the peptide, the surface and the solvent are larger than the total desorption energy, but opposite in sign while being of similar magnitude, so that they almost cancel out.