Wednesday, May 21, 2008

Effects of confinement

I have been feeling guilty at the slow pace with which I’ve been reading through a very nice preprint sent to me some time ago by Bob Eisenberg at Rush University Medical Center in Chicago, on the topic of how bubbles might act to gate ion channels. This brings together many recent ideas on how protein channels might be gated by evacuation of water from a hydrophobic pore – exactly the sort of thing discussed in the reviews by Gerhard Hummer et al. and Haiping Fang et al. in my previous post. The notion is that different channels have different mechanisms – mechanical, say, or voltage-depedent – to modulate the hydrophobicity of the channel and thus to cause an abrupt transition to a dewetted, capillary-evaporated state in which solutes are precluded from the channel. Eisenberg and colleagues present a general thermodynamic analysis of this process, and also offer the hypothesis that such a ‘bubble-induced’ mechanism might explain the anaesthetic effects of inert gases. The point is that the authors have now published the paper – or at least, what I assume is the same paper, as I’ve only so far seen the abstract – in Biophys. J. 94, 4282-4298 (2008) (paper here). Well worth looking at.

On the same general topic, Niharendu Choudhury in Mumbai has used MD simulations to look at how dewetting and capillary evaporation between two hydrophobic plates (close-packed paraffin monolayers) depends on the fine details of the plate structure (J. Phys. Chem. B 112, 6296-6300; 2008 – paper here). Specifically, he examines how the behaviour of the nano-confined water layer depends not only on plate separation but on the intermolecular distance in the paraffin plates, allowing a kind of ‘dilution’ of the hydrophobicity and solvent-surface dispersion forces. He finds that tuning this parameter can cause switches between wet, dry and intermittent wet/dry states, which might help to resolve differences seen in previous studies of this geometry. Moreover, the flickering formation and break-up of a water layer in the intermittent state can happen on nanosecond timescales, implying that simulations of protein hydration lasting only a few picoseconds may overlook important dynamical aspects of the problem.

Tobias Cramer at the University of Bologna and colleagues have looked at what one might consider the complementary problem: the spontaneous formation of a water bridge between two proximal surfaces (Langmuir 10.1021/la800220r – paper here). Their MD simulations examine how this process depends on an electric field across the gap, showing that there is a critical field strength at which the inhibitory influence of surface tension is overcome by electrostatic pressure. The focus here is mostly on scanning-probe-microscope-based experiments and technologies such as dip-pen nanolithography, but one can presumably imagine charging mechanisms for drawing water columns inside otherwise hydrophobic cavities in biomolecules. I have no idea whether such things are observed in nature…

There’s another take on nanoconfined water in a paper by Matthew Lane and colleagues at Sandia (J. M. D. Lane et al., Langmuir 24, 5209-5212; 2008 – paper here). They study the dynamics of a very thin film of water (submonolayer to bilayer) between two carboxyl-terminated alkanethiol self-assembled monolayers, using MD simulations. The diffusion coefficient of the water decreases as the layers become thinner, down to two orders of magnitude less than the bulk value, but the water remains liquid-like.

David Chandler, working with Adam Willard, has more on the role of solvent fluctuations in his dewetting model of hydrophobic assembly (discussed in earlier posts) in a paper in J. Phys. Chem. B 112, 6187-6192; 2008 (paper here). Fluctuations play a crucial role in the formation of a ‘vapour tunnel’ between two spherical hydrophobic particles, which draws them together. And in the same volume, Peter Rossky and colleagues expand on their work on the mechanism of protein cold denaturation, which I’ve also mentioned previously (C. F. Lopez et al., J. Phys. Chem. B 112, 5961-5967; 2008 – paper here).

Here’s another nice example of bound water playing a crucial role in enzyme function in a paper by Sason Shaik at the Hebrew University of Jerusalem and colleagues (Y. Wang et al., JACS 10.1021/ja711426y). They have looked at the mechanism by which cytochrome P450 StaP catalyses the formation of staurosporine, an antitumour agent, from chromopyrrolic acid. A critical step in this process is the abstraction of a proton from an N-H group on the substrate by an iron-oxo species in the enzyme. This seems to happen with the concerted assistance of two water molecules in the binding site: one shifts the proton onto a nearby histidine residue, and the other takes a proton from the other side of this residue’s side-chain and puts it on the iron-oxo group. Thus, the water molecules here form a hydrogen-bonded proton-relay network.

Rohit Pappu and colleagues at the University of St Louis in Missouri have an interesting study on the conformation of intrinsically disordered proteins, a class of protein that lack well defined 3D structures (H. T. Tran et al., JACS 10.1021/ja710446s – paper here). The common notion seems to be that because these proteins have sequences of low hydrophobicity, they are not tightly bound into compact structures by hydrophobic interactions. But IDPs are not totally random – they do have an ensemble of preferred conformational states. What creates them? The authors use simulations to conclude that these states are not dominated by specific intramolecular interactions in the polypeptide backbone, but from solvent-solute interactions, since water seems to be a generically poor solvent even for these low-hydrophobicity backbones.

More on hydration of poorly folded proteins comes from Supid Chakraborty and Sanjoy Bandyopadhyay at the Indian Institute of Technology in Kharagpur. They have used simulations to look at how the unfolding of the HP-36 subdomain of villin headpiece protein affects the dynamics of the hydration shell (J. Phys. Chem. B 112, 6500-6507; 2008 – paper here). Unfolding turns out to have a strong but quite complex effect on the rotational and translational motions of water in the hydration shell. It seems possible that there are knock-on effects: as one part of the protein unfolds, this can alter the structure and dynamics of hydration water around other segments in a cooperative manner. There’s clearly much more to be done on this interesting but under-investigated issue.

There’s more, as ever, to come when time permits…

Friday, May 9, 2008

Ions and water at interfaces: will they be understood by August?

There has been a lot of debate, some described in earlier posts, about the behaviour of hydroxide and hydronium ions at air-water and hydrophobic interfaces. Some claim the air-water interface is acidic, others that it is basic. So do protons or hydroxide ions get preferentially adsorbed at these interfaces? I’ve been sent a preprint (now published in J. Phys. Chem. C, doi:10.1021/jp800888b; paper here) by Robert Vácha, Ronen Zangi, Jan Engberts and Pavel Jungwirth that casts new light on the issue with simulations of hydroxide hydration near hydrophobic walls in aqueous KOH. They find that rigid walls create strong layering effects and a peak of hydroxide concentration about 5Å from the wall. But when the wall atoms are allowed to vibrate thermally, and when the dispersion interactions are weaker (more like the air-water interface), this structuring tends to get washed out, in some cases completely. Thus it seems one can’t generalize about hydroxide adsorption (or not) at a hydrophobic interface.

Another blow to ‘water structure effects’ at interfaces comes from a paper by Mischa Bonn and colleagues, in Huib Bakker’s group at FOM Amsterdam (M. Sovago et al., Phys. Rev. Lett. 100, 173901; 2008 – paper here). They have investigated the double-peaked vibrational sum-frequency generation (VSFG) spectrum of the O-H bonds of interfacial water in the hydrogen-bonded region, at a lipid-water interface. The two peaks have previously been interpreted as two distinct classes of hydrogen bond, ‘weak’ (ice-like) and ‘strong’ (water-like). But the FOM team, along with others from the Universities of Amsterdam and Utrecht, have evidence from isotope substitution experiments that the double peak is due to vibrational coupling between stretching and bending modes. They therefore conclude that the interfacial water is more homogeneous than has been thought.

How ions affect hydration and hydrophobic interactions is one of the thorniest problems in this field, and I won’t repeat myself by trying to summarize what has been said previously on the issue. Suffice to say that this question is intimately bound up with the notoriously puzzling ion-specific Hofmeister effects on solubility and aggregation of proteins. Neither am I going to make a poor attempt at summarizing the conclusions of a new investigation into these phenomena by Roland Netz and colleagues in Germany and Sweden (D. Horinek et al., Langmuir 24, 1271-1283; 2008 – paper here). Let’s just say that they have used single-molecule AFM experiments and MD simulations to delve into the ion-specific factors, free from complications of bubble nucleation and cavitation that may intervene for extended surfaces. The punchline is that “the most important factor determining ion-specific adsorption at hydrophobic surfaces can best be described as surface-modified ion hydration” – but you’d best read the paper to unpack that. A useful addition to a complicated story.

Lawrence Pratt and coworkers have a preprint exploring the role of dispersion forces on the potential of mean force between methane molecules in water. The find that these attractive methane-water interactions contribute a repulsive term to the pair potential (potential of mean force, pmf) between methanes. They also say that packing effects in the hydration shells make a dominant contribution to this pmf, but not in a way that can be interpreted with a perturbative approach – that is (if I’ve understood this properly), by treating the pmf as a perturbation expansion around this basic term.

Haiping Fang of the Shanghai Institute of Applied Physics and his coworkers have a very nice ‘topical review’ entitled ‘Dynamics of single-file water chains inside nanoscale channels: physics, biological significance and applications’ in J. Phys. D: Appl. Phys. 41, 103002 (2008). Some of the same issues, and some others, are addressed in a recent review by Gerhard Hummer and colleagues: J. C. Rasaiah et al., ‘Water in nonpolar confinement: from proteins to nanotubes and beyond’, Ann. Rev. Phys. Chem. 59, 713-740 (2008).

Finally, a free advert for a RSC Faraday Division discussion meeting on 27-29 August at Heriot-Watt University in Edinburgh entitled ‘Water: From Interfaces to the Bulk’ (see details here.). The announcement says that this meeting “plans to achieve a unification of views towards the goal of understanding the microscopic structure and behaviour of condensed phases of water at interfaces and progressing into the bulk.” I wish I could be there.

Friday, May 2, 2008

Life in the cold

[This is a slightly more considered summary of recent work on cryoprotectants, which appears as my Crucible column in the May issue of Chemistry World. The links to the papers discussed can be found in the previous blog entry below.]

When the going gets tough, the tough get sweet. There are many physiological responses to cold conditions, from goose pimples (useless for humans, handier for hairier beasts) to the famous antifreeze proteins of fish. But one of the common strategies for insects is to fill their cells with sugar. It’s still something of a mystery why this helps.

Cold poses diverse threats to life. Ice crystals in the body can simply rupture cell walls, which is why frozen strawberries thaw to a mush. And below about –20 degC protein molecules themselves start to unravel, a process called cold denaturation. That’s not well understood yet either, although a recent paper[1] suggests it involves weakening of the force between hydrophobic (water-repelling) parts of proteins that normally binds the folded form in place.

Sugars such as fructose and trehalose, as well as polyols such as glycerol and ethylene glycol, are manufactured seasonally by insects as cryoprotectants, just as we put antifreeze in our car radiators as winter draws near. Over winter, up to a fifth of the mass of some insects may consist of these substances. One consequence of cell fluid rich in sugar is simple depression of water’s freezing point. But that won’t get you very far into a bitter winter – typically, the depression will be only a few degrees, whereas the Arctic willow gall, flooded with glycerol, will survive temperatures of –66 degC. It’s thought that the cryoprotectants are doing something else here too.

Surviving the cold isn’t always about not getting frozen; sometimes there’s no avoiding it, and cryoprotectants then seem to act as freeze-tolerance rather than freeze-avoidance agents. It’s not only insects that do this: some frogs can survive being frozen solid if they fill their cells with glucose. The sugars and polyols seem to interact with cell water to protect delicate proteins and membranes – but no one is sure how.

Minoru Sakurai of the Tokyo Institute of Technology and his coworkers have shed some light on this through studies of the African midge Polypedilum vanderplanki[2]. They’ve studied not freezing as such, but an environmental stress more common in Africa which has similar consequences: dehydration. Dried larvae of the midge can enter a state called anhydrobiosis, in which they show no metabolic activity but can recover viability when water becomes available. They do this by generating trehalose.

There have been two suggestions for the protective mechanism: either the water is substituted by the sugar, or the sugar promotes the formation of a glassy cell matrix rather than ice crystals. The Japanese team thinks that in fact both are true. They find that the sugar forms hydrogen bonds with the lipids in cell membranes, replacing a shell of hydration water and preventing the membranes from becoming rigid. But the larvae also undergo a distinct glass transition as they are slowly dried. The glass is not pure trehalose, but is peppered with other components, such as proteins, that might help to disrupt crystallization.

How, though, does a shell of sugar or polyol protect a protein when water cannot? It seems that cryoprotectants can stabilize proteins against unfolding, but whether this comes from direct protein-sugar interactions or some kind of sugar-induced modification of water structure isn’t clear. Martina Havenith at the Ruhr University of Bochum and her colleagues recently reported signs that the latter might play a role[3]. Using terahertz spectroscopy, they found that the dynamics of water molecules are disturbed a remarkably long distance away from dissolved sugars – up to about 5-7 Å for trehalose and lactose. These perturbations are stronger and longer-ranged for disaccharides than for the monosaccharide glucose, which would support the notion that cryoprotection (which disacchardides do better) is tied up with the sugar’s ability to slow down the water motions and promote a pseudo-glassy state.

Findings by Giovanni Strambini and coworkers at the Consiglio Nazionale delle Ricerche in Pisa, Italy, could be seen to lend support to this idea. The Italian team have asked how cryoprotectants do their job if ice actually begins to form. They used fluorescence spectroscopy to study the stabilization of a protein called azurin by sugars and polyols (sucrose, trehalose, sorbitol, glycerol) in ice-water mixtures[4]. It seems none of these molecules offers strong protection against ice formation, although trehalose ‘tries hardest’: as ice appears, the protein is increasingly prone to unfold. So the cryoprotectants don’t make the native protein significantly more thermodynamically stable. Instead, the researchers think that they somehow cajole the protein to stay folded in the liquid until the whole system becomes a sluggish glass and unfolding is then simply too slow – a kinetic rather than thermodynamic effect.

So there is an emerging picture, albeit a complex one. The cryoprotectants could have a dual role. First they remodel biomolecular hydration shells, retarding water and maybe suppressing the loss of crucial protein-folding forces. Then they eventually promote the formation of a glassy matrix rather than an icy one, arresting the biomolecular structures in recoverable suspended animation. That’s clever work for a spoonful of sugar.


1. C. L. Dias et al., Phys. Rev. Lett. 100, 118101 (2008).
2. M. Sakurai et al., Proc. Natl Acad. Sci. USA 105, 5093-5098 (2008).
3. M. Heyden et al., J. Am. Chem. Soc. doi:10.1021/ja0781083.
4. G. B. Strambini et al., J. Phys. Chem. B 112, 4372-4380 (2008).