Chem Rev article is now online

My review article on water in cell biology has now been published online in Chemical Reviews – it’s available here. Some minor corrections in the first five pages were omitted (I’m hoping these might yet be fixed), but such is life. Hope it’s useful.

Spherical water

More on the issue of urea denaturation of proteins comes from Paul Cremer and colleagues at Texas A&M (Chen et al., JACS 129, 15104; 2007). They have used vibrational sum frequency spectroscopy to look at the orientation of urea molecules at the surface of bovine serum albumin. They find that the orientation depends on the surface charge on the protein: it flips from pointing the amines towards the surface at high pH (negative surface charge) to pointing the carbonyls at low pH (positive surface charge). That’s interesting. But the authors’ argument that this supports an indirect model for denaturation, where the urea primarily acts via a restructuring of the hydration sphere rather than direct surface bonding, seems rather vague and not at all obvious at this point.

Sergey Buldyrev at Yeshiva University and coworkers (Kumar, Debenedetti, Rossky, Stanley) have found water-like behaviour in a rather un-water-like system (PNAS, doi:10.1073/pnas.0708427104). They’ve looked at a fluid of spherically symmetric particles with two interaction length scales – a hard core and a soft repulsive ‘ramp’ – and see water-like anomalies such as expansion on cooling, as well as cold-induced ‘denaturation’ of a hard-sphere polymer chain. This suggests that this dual-scale characteristic might be the most fundamental feature that gives water its ‘uniqueness’.

Ivan Brovchenko’s paper on water percolation effects in DNA polymorphism, which I mentioned several weeks back, has now been published in JACS: doi:10.1021/ja0732882.

More water tuning, and out with the structure-breakers

I couldn’t hope for a better illustration of how biomolecular hydration can be used to fine-tune function than that provided in a paper in Science by Keith Hodgson at Stanford and coworkers (30 November, Vol. 318, p.1464). They have used sulphur XAS to look at the iron-sulphur sites in the high-potential iron-sulphur protein (HiPIP) and in ferredoxin (Fd). Both have much the same iron-sulphur cluster, yet that in HiPIP gets oxidized and that in Fd gets reduced at physiological potentials. The authors conclude that this difference is due to different degrees of Fe-S covalency, which in turn depends on the hydration state. In other words, function here has in some sense rather little to do with the protein’s primary structure, but depends instead on the hydration environment.

Maria Ricci in Rome and her colleagues have a great paper in J. Phys. Chem. (Vol. 111, p.13570) on the structure of NaCl and KCl solutions as deduced by neutron scattering. They find that K ions have more orientationally disordered hydration shells than Na ions, while Cl ions tend to form H-bonded bridges between waters. In both cases the H-bonded structure of water is significantly disrupted, but not in a way that yields to any simplistic description as ‘structure-making’ or ‘structure-breaking’. As a result, the authors say, those old ideas “are not helpful in understanding how these ions interact with water at the molecular level.” Hear, hear. Let’s hope that message gets out.

There’s what looks to be a very nice paper in Biophys. J. (Vol. 93, p.4116) from Bryan Patel, Pablo Debenedetti, Frank Stillinger and Peter Rossky on the way hydrophobic hydration acts to cause protein denaturation at low and high temps and at high pressure. I’ve only seen the abstract of this so far, but the comment that “an explicit treatment of hydrophobic hydration is sufficient to produce cold, pressure, and thermal denaturation” implies that this paper covers a lot of important ground in this controversial area.

In a paper in J. Phys. Chem. B (doi:10.1021/jp077110d), Julio Martinez and Pieter Stroeve suggests that they can resolve previously discrepant findings about the nature of the interface between water and a hydrophobic surface. They have used surface plasmon resonance to study this interface for a self-assembled monolayer, and report that it evolves slowly: at first, nanobubbles are formed, but these disappear after about 10 minutes. Equilibrium is not reached, however, until about 30 hours later. In this equilibrium state, there is apparently an ‘organic layer’ at the interface, by which they seem to mean a film of organic contaminants. Something like that was reported by Evans et al. in Physica A 339, 101 (2004). I’m not sure this will be the final word, or quite what it implies for hydrophobic interactions, but it does seem to offer a reason why previous results have differed.

There’s going to be more to be said too about the issue of ions at the air-water or hydrophobe-water surface. Pavel Jungwirth and coworkers made a small splash earlier this year with a PNAS paper that claimed to find surface acidification. Greg Voth predicted that some time ago (se below, ‘Acid on top’, http://waterinbiology.blogspot.com/2007_04_01_archive.html). But others insist that hydroxide ions preferentially segregate at the surface. Jan Engberts now tells me that “Jungwirth has now performed MD simulations, with the help of my previous post-doc Ronen Zangi, now at Columbia [with Bruce Berne]. And they largely reproduce our previous results, i.e. hydroxide binding....The idea is now to write a joint paper. I am not sure what comes out of it.” Watch this space… And meanwhile, Dominic Horinek and Roland Netz have a simulation paper in Phys. Rev. Lett. (Vol. 99, 226104) which reports that large, polarisable halide ions are adsorbed preferentially at the surface of a hydrophobic SAM.

Sticking hydrophobes with salts and smallness

Do Hofmeister effects after all depend on altering ‘water structure’? Frankly, I doubt it. But a suggestive case is made in a paper by Andrew Thomas and Adrian Elcock (JACS doi:10.1021/ja073097z). Their MD simulations of various salt solutions show that changes in water-water hydrogen bonding appear to be correlated with experimental solubility data for hydrophobic solutes. Strongly salting-out salts, for instance, cause significant decreases in the water-water hydrogen-bonding fraction. Lithium ions, previously considered anomalous in their salting-out behaviour, form linear ionic chains, with correspondingly unusual hydration structures. But is all this behaviour seen in neutron-scattering studies of salt solutions? I don’t recall that it is. In any event, Thomas and Elcock also find that for simulations that include hydrophobes, an increase in hydrophobic association for certain hydrophobes and salts also correlates with solubility data. There’s a way to go yet before we understand all this.

Relevant to this paper is an experimental study by Jared Smith, Rich Saykally and Phillip Geissler (JACS 129, 13847; 2007) on the effects of dissolved halide ions on hydrogen bonding in water. In contrast to the old ideas about structure-making/breaking, they find that the effects on Raman and IR vibrational spectra can be explained by the action of the ions’ electric fields on adjacent water molecules, and that H-bond strengths are altered very little beyond the first hydration shell. In other words, the H-bond network seems rather robust to such perturbations.

Hydrophobic association in pure water is studied by K. G. Ayappa and colleagues at the Indian Institute of Science in Bangalore (Langmuir doi:10.1021/la7022902). They consider the effects of nanoconfinement on the interaction, looking at 2.82-nm diameter water droplets in reverse micelles. They find that the attraction is enhanced by the confinement, which they explain by the lack of sufficient water to solvate and stabilize the solvent-separated solutes. Plausible? I guess so – after all, hydration of lone hydrophobes is thermodynamically favourable. Dave Thirumalai has considered this issue recently (JACS 128, 13490; 2006) – I must remind myself of what he found…

Protein-water coupling: confirmations and complications

With far too much to catch up with here, I shall do little more than list things that have crossed my radar screen. Lots happening, all interesting…

Alla Oleinikova, Nikolai Smolin, and Ivan Brovchenko have a paper in Biophys. J. (93, 2986) entitled “Influence of Water Clustering on the Dynamics of Hydration Water at the Surface of a Lysozyme”, in which they use MD simulations to look at the coupling of water and protein dynamics as the degree of hydration changes. In line with their earlier work, they see maximal dynamical coupling when the water coverage corresponds to a percolating water network on the protein surface.

Ivan and Alla have also told me about their forthcoming book, Interfacial and Confined Water, to be published by Elsevier, which will look at water’s behaviour at hydrophilic and hydrophobic surfaces in general but with clearly a pretty strong focus on biomolecules, including these ideas about percolation transitions in the hydrogen-bonded network.

The hydration dynamics at a protein surface are also the topic of a paper from Dongping Zhong and colleagues at Ohio State University (PNAS doi:10.1073/pnas.0707647104). They have used ultrafast spectroscopy to map out the hydration dynamics from place to place on the surface of various mutants of sperm whale myoglobin, and find two distinct dynamical regimes: one with dynamical timescales of 1-8 ps, the other with around 20-200 ps. These regimes are strongly correlated with the protein’s structure and composition, confirming the intimate relationship between hydration dynamics and protein fluctuations.

But at the same time, this story gets more complex. Martin Weik has sent me a forthcoming paper to be published in PNAS (doi:10.1073/pnas.0706566104) called “Coupling of protein and hydration-water dynamics in biological membranes”. Here they use inelastic neutron scattering and MD simulations to look at the relationship between water dynamics and fluctuations of lipids and bacteriorhodopsin in the purple membrane between 120 and 260 K. They find that the two seem to be decoupled, at least below 260 K, in contrast to the situation for soluble proteins and their hydration layers. In other words, there is no coupled ‘glass-like’ transition of the water and membrane protein: the onset of water motion as the temperature is raised through 200 K does not coincide with a dynamical transition of bR. That adds a whole new layer of complexity to the ongoing story of protein-water dynamics: membranes change the game.

Time to change the subject, then. The hydration of DNA tends to get far less attention than that of proteins, but evidently has interesting stories attached. It seems fairly clear now that the regular double helix depends on the presence of water, though that tends to be glossed over in biochemical texts. Hermann Gaub and colleagues have now made that point in a very forceful manner (JACS doi:10.1021/ja074776c). They have used an AFM tip attached to one strand to drag a length of double-stranded DNA from water into a poor (nonpolar) solvent, octane - whereupon the ds-DNA unzips spontaneously. This happens too in MD simulations. That, the authors say, might be exploited by helicases, which need only force the DNA into a hydrophobic binding pocket to make it unwind. A lovely and striking result.

Return of the iceberg model?

A paper by Huib Bakker and Yves Rezus in Phys. Rev. Lett. (vol. 99, 148301; 5 Oct.) seems bound to stir up some debate. The work seems nice: an ultrafast IR spectroscopic study of water motions in the hydration spheres of some small organic molecules, which apparently indicates that the (four or so) waters hydrating the methyl groups are rotationally retarded by a factor of at least 4-5 relative to the bulk, while the other waters in the hydration sphere are barely affected. Bakker has used this technique extensively, and one would imagine the results are reliable. Indeed, they seem very much in line with what has been reported previously, for example from NMR studies.

They interpret the slowing as being due to steric hindrance of the breaking of hydrogen bonds via a five-coordinate species – somewhat akin, if I remember rightly, to the kind of slowing down of SN2 substitution reactions in organic chemistry when they are similarly sterically blocked.

What is curious is that the discussion is framed in the context of the Frank & Evans hypothesis from 1945 of an ‘ice-like’ hydration sphere for hydrophobic groups (H. S. Frank & M. W. Evans, J. Chem. Phys. 13, 507; 1945). The paper itself seems to imply that the findings validate this picture – and as a consequence, support the 1959 idea of Walter Kauzmann of an entropic basis for the hydrophobic interaction.

The problem is that that idea seems inconsistent with just about all previous experimental evidence (see, for example, Blokzijl and Engberts, Angew. Chem. Int. Ed. 32, 1545; 1993). And I can’t for the life of me see why a factor of several-fold slowing of rotation should be equated with ‘immobilization’ of the water. Yet this is how the work seems to be getting sold by the APS. (See http://focus.aps.org/story/v20/st11; note in particular, “Biophysicist Kim Sharp of the University of Pennsylvania considers this the first direct observation of the iceberg model, thus completing a long history of trying to confirm this theory” – and the statement in the paper itself that “Our results provide a molecular picture of these icebergs”. Gulp.) Given how entrenched the Kauzmann model has become, without good reason, it seems unfortunate that it as apparently going to receive further support from this work, without any real justification that I can see.

A(nother) word on urea

I discovered at the 2007 Halophiles meeting at the University of Essex earlier this month that the mechanism of protein denaturation by urea is still a matter of debate. That’s not, perhaps, terribly surprising in view of the fact that even the hydration structure of urea itself is not certain, as earlier posts have mentioned. Jose Manuel Hermida-Ramon at the University of Vigo in Spain and coworkers add a contribution to this debate in J. Phys. Chem. B [doi:10.1021/jp073579x]. They use quantum-chemical calculations to deduce the structure of the hydrated urea molecule, and say that it is ill-defined: the molecule is very floppy, because the transition from a planar to a non-planar structure has an activation energy comparable to the room-temperature thermal energy. However, they say that urea might adopt a fixed, or less flexible, structure, as it approaches a protein surface.

Wayne Bolen and colleagues at the University of Texas Medical Branch at Galveston have attempted to tease out the ways that urea interacts with peptide residues when this happens [M. Auton et al., PNAS doi:10.1073/pnas.0706251104]. Using thermodynamic data, they say that, contrary to some previous views, the key interactions are not with nonpolar side chains, but involve the peptide backbone itself, and that these latter interactions are what drives denaturation. No doubt we’ll be hearing more about this issue.

The question of dewetting of protein surfaces in folding and aggregation also rumbles on. Following on from the Lum/Chandler/Weeks idea of dewetting of large hydrophobes and a consequent crossover length in the mechanism of hydrophobic attraction [K. Lum et al., J. Phys. Chem. B 103, 4570; 1999; D. Chandler, Nature 437, 640; 2005], Jeremy Smith at Heidelberg and colleagues have looked at whether there is ‘dewetting’ around hydrophobic residues of smaller peptides [I. Daidone et al., PNAS doi:10.1073/pnas.0701401104]. They say that for a 14-residue beta-hairpin peptide, conformers that expose significant amounts of hydrophobic surface have a lower hydration density than those that don’t, and that as a consequence, “dehydration-driven solvent exposure of hydrophobic surfaces may be a significant factor determining peptide conformational equilibria.” Which looks fine as far as it goes, but I can’t obviously see if this addresses the question of whether there is an abrupt, cooperative drying transition during folding, as seemed to be a central feature of the LCW model…

Collapse and cooperation in water

I seem to have missed the recent paper by David Chandler and colleagues on collapse of a hydrophobic polymer chain within a ‘coarse-grained’ model of a water solvent [PNAS 104, 14559]. I was alerted to it by the commentary in the forthcoming issue of PNAS by Gerhard Hummer [doi:10.1073/pnas.0706633104]. David’s paper provides support for his suggestion, with ten Wolde, that hydrophobic polymer collapse happens via a dewetting transition [PNAS 99, 6539; 2002]. In the new simulations, expulsion of water through collective motions is the rate-limiting step of the collapse, and moreover it is the work performed on the solvent in this process that supplies the free-energy barrier – that is, dewetting doesn’t passively accompany the collapse, but drives it.
I’ve talked about this idea a fair bit in previous posts in relation to protein folding and aggregation: Bruce Berne’ studies have suggested that dewetting transitions can happen in this context, but are not the general rule. There doesn’t seem to be any obvious inconsistency between these findings: David is looking simply at hydrophobic chains, whereas it seems that only a few polar groups in the chain, as are generally found in proteins, can be sufficient to suppress dewetting. That, at this point, seems to be the story.

There are two nice illustrations of the roles of hydration water in protein function in the latest ASAP section of JACS. Mario Rivera at the University of Kansas and coworkers have used NMR relaxation to look at the role of a hydrogen-bonded network of waters in the function of a bacterial heme oxygenase [J. C. Rodriguez et al., doi:10.1021/ja072405q]. They find that the network serves three roles. First, it conducts protons to the iron-dioxygen complex during catalysis. Second, it propagates changes in electronic structure at the active site during the course of the reaction to remote parts of the polypeptide. Third, it modulates the conformational freedom of the enzyme, allowing it to accommodate and adapt to the changes in conformation required during the catalytic process. If the hydrogen-bonded network is disrupted in a mutant form, the necessary coordination in dynamics of different parts of the protein is lost and the motions become almost globally chaotic, lowering the efficiency of the enzyme significantly. In short, the enzymatic process simply ‘makes no sense’ without the aid of the waters. As the authors put it, “The information needed to tune the dynamic freedom of the polypeptide is communicated from the active site to the polypeptide via the hydrogen-bonding network.” That’s a wonderful example of the delicate fine-tuning of structure and dynamics that these hydration structures can offer.

Second, Vicent Moliner at the Universitat Jaume I in Castello and colleagues use simulations to confirm the idea that long-distance electron transfer between the metal sites in a dicopper enzyme (peptidylglycine alpha-hydroxylating monooxygenase) is mediated by a bridge of hydrogen-bonded water molecules and peptide residues [de la Lande et al., doi:10.1021.ja070329l] This has been suggested for some cytochromes too.

On the surface

I’ve been sent an advance copy of a nice review article on ‘water at solid surfaces’ by Marco Maccarini at Heidelberg. It will appear in the September issue of Biointerphases (a journal of the American Vacuum Society), and does a thorough job of reviewing recent work on the nature of water at hydrophobic and hydrophilic surfaces as revealed by X-ray and neutron reflectivity, SHG and simulation studies.

And on that same topic, there’s a paper in Physical Review Letters (99, 078302) reporting the nature of water at the interface with a phospholipid membrane, based on 2D IR spectroscopy. The authors conclude that there are three types of hydrogen-bonding motif: water molecules bound to zero, one or two OH groups on the lipids, in the relative proportions of about 8:52:40. Can’t immediately see any data about the lifetimes.

A bad memory

[Please forgive the double posting here - it is also on my regular blog www.philipball.blogspot.com - and also forgive its informality, for the same reason. But it seemed relevant to put this up here too.]
I have just read all the papers on ‘the memory of water’ published in a special issue of the journal Homeopathy, which will be released in print on 10 August. Well, someone had to do it. I rather fear that my response, detailed below, will potentially make some enemies of people with whom I’ve been on friendly terms. I hope not, however. I hope they will respect my right to present my views as much as I do theirs to present theirs. But I felt my patience being eroded as I waded through this stuff. Might we at least put to rest now the tedious martyred rhetoric about ‘scientific heresy’, which, from years of unfortunate experience, I can testify to being the badge of the crank? I once tried to persuade Jacques Benveniste of how inappropriate it was to portray a maverick like John Maddox as a pillar of the scientific establishment – but he wouldn’t have it, I suppose because that would have undermined his own platform. Ah well, here’s the piece, a much shortened version of which will appear in my Crucible column in the September issue of Chemistry World.

************************************************************

I met Jacques Benveniste in 2004, shortly before he died. He had tremendous charm and charisma, and I rather liked him. But I felt then, and still feel now, that in ‘discovering’ the so-called memory of water he lost his way as a scientist and was sucked into a black hole of pseudoscience that was just waiting for someone like him to come along.

This particular hole is, of course, homeopathy. In 1988, Benveniste published a paper in Nature that seemed to offer an explanation for how homeopathic remedies could retain their biological activity even after being diluted so much that not a single molecule of the original ‘active’ ingredients remains [1]. It is common for homeopathic remedies to have undergone up to 200 tenfold dilutions of the original ‘mother tincture’, which is quite sufficient to wash away even the awesome magnitude of Avogadro’s constant.

Benveniste and his coworkers studied the effect of dilution of an antibody that stimulates human immune cells called basophils to release histamine – a response that can provoke an allergic reaction. In effect, the antibody mimics an allergen. The researchers reported that the antibody retains its ability to provoke this response even when diluted by 10**60 – and, even more oddly, that this activity rises and falls more or less periodically with increasing dilution.

The paper’s publication in Nature inevitably sparked a huge controversy, which turned into a media circus when Nature’s then editor John Maddox led an investigation into Benveniste’s laboratory techniques. Several laboratories tried subsequently to repeat the experiment, but never with unambiguous results. The experiment proved irreproducible, and came to be seen as a classic example of what US chemist Irving Langmuir christened ‘pathological science’. (The details are discussed in my book on water [2], or you can read Michel Schiff’s book [3] for a deeply partisan view from the Benveniste camp.)

Benveniste remained convinced of his results, however, and continued working on them in a privately funded lab. He eventually claimed that he could ‘programme’ specific biological activity into pure water using electromagnetic radiation. He predicted a forthcoming age of ‘digital biology’, in which the electromagnetic signatures of proteins and other biological agents would be digitally recorded and programmed into water from information sent down phone lines.

Homeopaths have persistently cited Benveniste’s results as evidence that their treatments do not necessarily lack scientific credibility. Such claims have now culminated in a special issue of the journal Homeopathy [4] that presents a dozen scientific papers on the ‘memory of water.’

In at least one sense, this volume is valuable. The memory of water is an idea that refuses to go away, and so it is good to have collected together all of the major strands of work that purport to explain or demonstrate it. The papers report some intriguing and puzzling experimental results that deserve further attention. Moreover, the issue does not duck criticism, including a paper from renowned water expert José Teixeira of CEA Saclay in France that expresses the sceptic’s viewpoint. Teixeira points out that any explanation based on the behaviour of pure water “is totally incompatible with our present knowledge of liquid water.”

But perhaps the true value of the collection is that it exposes this field as an intellectual shambles. Aware that I might hereby be making enemies of some I have considered friends, I have to say that the cavalier way in which ‘evidence’ is marshalled and hypotheses are proposed with disregard for the conventions of scientific rigour shocked even me – and I have been following this stuff for far too long.

Trying to explain homeopathy through some kind of aqueous ‘memory’ effect has plenty of problems created by the traditions of the field itself, in which ‘remedies’ are prepared by serial dilution and vigorous shaking, called succussion. For example, it is necessary not only that the memory exists but that it is amplified during dilution. In his overview paper, guest editor Martin Chaplin, a chemist at South Bank University in London whose web site on water is a mine of valuable information, points to the surprising recent observation that some molecules form clusters of increasing size as they get more dilute. But this, as he admits, would imply that most homeopathic solutions would be totally inactive, and only a tiny handful would be potent.

Another problem, pointed out by David Anick of the Harvard Medical School and John Ives of the Samueli Institute for Information Biology in Virginia, is that if we are to suppose the ‘memory’ to be somehow encoded in water’s structure, then we must accept that there should be many thousands of such stable structures, each accounting for a specific remedy – for several thousand distinct remedies are marketed by homeopathic companies, each allegedly distinct in its action.

Yet another difficulty, seldom admitted by homeopaths, is that the dilutions of the mother tincture must allegedly be made by factors of ten and not any other amount. This is not mentioned in the papers here, presumably because it is too absurd even for these inventive minds to find an explanation. A related issue that is addressed by Anick is the tradition of using only certain dilution factors, such as 10**6, 10**12, 10**30 and 10**200. He offers a mathematical model for why this should be so that masquerades as an explanation but is in fact tantamount to a refutation: “it would be inconceivable”, he says, “that one number sequence would work in an ideal manner for every mother tincture.” Still, he concludes, the convention might be ‘good enough’. So why not perhaps test if it makes any difference at all?

One of the challenges in assessing these claims is that they tend to play fast and loose with original sources, which obliges you to do a certain amount of detective work. For example, Chaplin states that the ability of enzymes to ‘remember’ the pH of their solvent even when the water is replaced by a non-aqueous solvent implies that the hydrogen ions seem to have an effect in their absence, “contrary to common sense at the simplistic level.” But the paper from 1988 in which this claim is made [5] explains without great ceremony that the ionizable groups in the enzyme simply retain their same ionization state when withdrawn from the aqueous solvent and placed in media that lack the capacity to alter it. There’s no mysterious ‘memory’ here.

Similarly, Chaplin’s comment that “nanoparticles may act in combination with nanobubbles to cause considerable ordering within the solution, thus indicating the possibility of solutions forming large-scale coherent domains [in water]” is supported by a (mis-)citation to a paper that proposes, without evidence, the generally discredited idea of ‘ice-like’ ordering of water around hydrophobic surfaces.

One of the hypotheses for water’s ‘memory’, worked out in some detail by Anick and Ives, invokes the dissolution of silicate anions from the glass walls of the vessel used for dilution and succussion, followed by polymerization of these ions into a robust nanostructured particle around the template of the active ingredient initially present. Certainly, silicate does get added, in minute quantities, to water held in glass (this seemed to be one of the possible explanations for another piece of water pathological science, polywater [6]). But how to progress beyond there, particularly when such a dilute solution favours hydrolysis of polysilicates over their condensation?

Well, say Anick and Ives, there are plenty of examples of silicate solutions being templated by solutes. That’s how ordered mesoporous forms of silica are synthesized in the presence of surfactants, which aggregate into micelles around which the silica condenses [7]. This, then, wraps up that particular part of the problem.

But it does nothing of the sort. This templating has been seen only at high silicate concentrations. It happens when the template is positively charged, complementary to the charge on the silicate ions. The templating gives a crude cast, very different from a biologically active replica of an enzyme or an organic molecule. Indeed, why on earth would a ‘negative’ cast act like the ‘positive’ mold anyway? The template is in general encapsulated by the silica, and so doesn’t act as a catalyst for the formation of many replicas. And for this idea to work, the polysilicate structure has to be capable of reproducing itself once the template has been diluted away – and at just the right level of replicating efficiency to keep its concentration roughly constant on each dilution.

The last of these requirements elicits the greatest degree of fantastical invention from the authors: during the momentary high pressures caused by succussion, the silicate particles act as templates that impose a particular clathrate structure on water, which then itself acts as a template for the formation of identical silicate particles, all in the instant before water returns to atmospheric pressure. (Elsewhere the authors announce that “equilibrium of dissolved [silicate] monomers with a condensed silica phase can take months to establish.”) None of this is meanwhile supported by the slightest experimental evidence; the section labelled ‘Experiments to test the silica hypothesis’ instead describes experiments that could be done.

Another prominent hypothesis for water’s memory draws on work published in 1988 by Italian physicists Giuliano Preparata and Emilio Del Guidice [8]. They claimed that water molecules can form long-ranged ‘quantum coherent domains’ by quantum entanglement, a phenomenon that makes the properties of quantum particles co-dependent over long ranges. Entanglement certainly exists, and it does do some weird stuff – it forms the basis of quantum computing, for example. But can it make water organize itself into microscopic or even macroscopic information-bearing domains? Well, these ‘quantum coherent domains’ have never been observed, and the theory is now widely disregarded. All the same, this idea has become the deus ex machina of pathological water science, a sure sign that the researchers who invoke it have absolutely no idea what is going on in their experiments (although one says such things at one’s peril, since these researchers demonstrated a litigious tendency when their theory was criticized in connection with cold fusion).

Such quantum effects on water’s memory are purportedly discussed in the special issue by Otto Weingärtner of Dr Reckeweg & Co. in Bensheim, Germany – although the paper leaves us none the wiser, for it contains neither experiments nor theory that demonstrate any connection with water. The role of entanglement is made more explicit by Lionel Milgrom of Imperial College in London, who says that “the homeopathic process is regarded as a set of non-commuting complementary observations made by the practitioner… Patient, practitioner, and remedy comprise a three-way entangled therapeutic entity, so that attempting to isolate any of them ‘collapses’ the entangled state.” In other words, this notion is not really about quantum mechanics at all, but quantum mysticism.

Benveniste’s long-term collaborator Yolène Thomas of the Institut Andre Lwoff in Villejuif argues, reasonably enough, that in the end experiment, not theory, should be the arbiter. And at face value, the ‘digital biology’ experiments that she reports are deeply puzzling. She claims that Benveniste and his collaborators accumulated many examples of biological responses being triggered by the digitized radiofrequency ‘fingerprints’ of molecular substances – for example, tumour growth being inhibited by the ‘Taxol signal’, the lac operon genetic switch of bacteria being flipped by the signal from the correct enantiomeric form of arabinose, and vascular dilation in a guinea pig heart being triggered by the signal from the classic vasodilator acetylcholine. What should one make of this? Well, first, it is not clear why it has anything to do with the ‘memory of water’, nor with homeopathy. But second, I can’t help thinking that these experiments, however sincere, have an element of bad faith about them. If you truly believe that you can communicate molecular-recognition information by electromagnetic means, there is no reason whatsoever to study the effect using biological systems as complex as whole cells, let alone whole hearts. Let’s see it work for a simple enzymatic reaction, or better still, an inorganic catalyst, where there is far less scope for experimental artefacts. It is hard to imagine any reason why such experiments have not been attempted, except for the reason that success or failure would be less ambiguous.

What emerges from these papers is an insight into the strategy adopted more or less across the board by those sympathetic to the memory of water. They begin with the truism that it is ‘unscientific’ to simply dismiss an effect a priori because it seems to violate scientific laws. They cite papers which purportedly show effects suggestive of a ‘memory’, but which often on close inspection do nothing of the kind. They weave a web from superficially puzzling but deeply inconclusive experiments and ‘plausibility arguments’ that dissolve the moment you start to think about them, before concluding with the humble suggestion that of course all this doesn’t provide definitive evidence but proves there is something worth further study.

One has to conclude, after reading this special issue, that you can find an ‘explanation’ at this level for water’s memory from just about any physical phenomenon you care to imagine – dissipative non-equilibrium structures, nanobubbles, epitaxial ordering, gel-like thixotropy, oxygen free radical reactions… In each case the argument leaps from vague experiments (if any at all) to sweeping conclusions that typically take no account whatsoever of what is known with confidence about water’s molecular-scale structure, and which rarely address themselves even to any specific aspect of homeopathic practice. The tiresome consequence is that dissecting the idea of the memory of water is like battling the many-headed Hydra, knowing that as soon as you lop off one head, another will sprout.

In his original paper in Nature, Jacques Benveniste offered a hypothesis for how the memory effect works: “specific information must have been transmitted during the dilution/shaking process. Water could act as a template for the [antibody] molecule, for example by an infinite hydrogen-bonded network or electric and magnetic fields.” Read these sentences carefully and you will perhaps decide that Benveniste missed his calling as a post-modernist disciple of his compatriot Jacques Derrida. It has no objective meaning that I can discern. It sounds like science, but only because it copies the contours of scientific prose. This, I would submit, is a fair metaphor for the state of ‘water memory’ studies today.

I once read a book supposedly about the philosophy of religion which was in fact an attempt to make a logical case for God’s existence. Having stepped through all of the traditional arguments – the ontological, the argument from design and so forth – the author admitted that all of them had significant flaws, but concluded that collectively they made a persuasive case. This group of papers is similar, implying that a large enough number of flimsy arguments add up to a single strong one. It leaves me feeling about homeopathy much as I do about religion: those who find it genuinely helpful are right to use it, but they shouldn’t try to use scientific reason to support their decision.


1. E. Davenas et al., Nature 333, 816 (1988).
2. P. Ball, H2O: A Biography of Water (Weidenfeld & Nicolson, 1999).
3. M. Schiff, The Memory of Water (Thorsons, 1995).
4. Homeopathy 96, 141-226 (2007).
5. A. Zaks & A. Klibanov, J. Biol. Chem. 263, 3194 (1988).
6. F. Franks, Polywater (MIT Press, Cambridge, MA, 1981).
7. C. T. Kresge et al., Nature 359, 710 (1992).
8. E. Del Guidice et al. Phys. Rev. Lett. 61, 1085 (1988).

More of the same

Two papers provide additional evidence of perturbations to water’s structure and properties at interfaces. The long saga about density depletion at hydrophobic surfaces now seems to be settling down in favour of the position that there is indeed a degree of depletion when the surfaces are strongly hydrophobic, but only over angstrom distances. That is supported by the Monte Carlo simulations of Jiri Janacek and Roland Netz (Langmuir 23, 8417; 2007), who see depletion layers 1.5-2 Å thick at ordered hydrocarbon surfaces with contact angles of 110-130 degrees.

And the idea that water confined at the nanoscale between surfaces has greatly enhanced viscosity (see, for example, Li et al., Phys. Rev. B 75, 115415; 2007) is supported in further experiments by Tai-De Li and Elisa Riedo at Georgia Tech, who have investigated water’s nonlinear viscoelastic behaviour using the AFM (www.arxiv.org/abs/ 0707.2521).

A proton switch in GFP

Emission from green fluorescent protein (GFP) is one of the most widely used methods of molecular marking in cell biology, since GFP can be prepared as a fusion protein with just about any other gene product. But the fluorescence behaviour is curious and hasn’t been fully explained. In particular, it shows a t**-3/2 time dependence in the long-time tail at room temperature, but switches to a t**-1/2 dependence below 230 K. Fluorescence involves a photoexcited proton transfer from the chromophore, which is thought to occur along a hydrogen-bonded chain involving various residues and bound water molecules. Noam Agmon has modelled this process, considering the ‘proton wire’ to be rather longer than is normally thought and to have within it a switch at a threonine residue (Thr203) with a large activation energy for proton migration (J. Phys. Chem. B 111, 7870; 2007). The rapid migration of the proton is held up at this switch point for typically 300 ps at room temperature. This model can explain both the t**-3/2 behaviour at room temperature and the changeover to different asymptotics at 230 K. Here’s another example of water and hydrogen-bonding residues collaborating to engineer biological function.

Proteins that dry in a flash

Do proteins aggregate and fold in an abrupt ‘dewetting’ transition that expels water from between hydrophobic surfaces, or is the water squeezed out more gradually? The former idea has been popularised by Lum, Weeks and Chandler (J. Phys. Chem. B 103, 4570; 1999), who argued that this drying transition should be expected for surfaces of around 1 nm or more in at least one dimension. But observations and simulations of protein aggregation and folding haven’t generally supported it (see, for example, Zhou et al, Science 305, 1605; 2004). Yet Bruce Berne and his colleagues (who conducted that study in Science) have found that the tetrameric channel-forming protein melittin does seem to show a dewetting transition (Liu et al., Nature 437, 159; 2005). Is that a rarity, even a unique case, or might other proteins also exhibit dewetting? Berne and co. have performed a survey of the protein data bank to search for other structures that might show similar behaviour (Hua et al., J. Phys. Chem. B, 10.1021/jp0704923). The message is that dewetting is rare, but does happen in a few other cases too: the authors find several other examples of multi-domain proteins that display it in the final stages of folding. Specifically, they identify two two-domain proteins six dimers and three tetramers that behave this way. It seems that any significant number of polar residues in the hydrophobic core (which is common) is generally enough to suppress dewetting. Using the same tools, however, Berne and colleagues find preliminary evidence that dewetting may also sometimes play a role in ligand binding.

What proteins do to water

Why does protein hydration water display anomalous dynamics? There is a huge literature on this, particularly in terms of quantifying the effect. Clearly the effect is heterogeneous and complex, but there is some reason to suppose that in general hydration water exhibits anomalous diffusion, the translational motion following a t**0.6 time dependence rather than Brownian t**0.5.

Francesco Pizzitutti at Saclay, Peter Rossky at Texas at Austin, and their coworkers have tried to figure out what is going on using MD simulations (J. Phys. Chem. B 111, 7584; 2007). They consider two aspects of the problem: the effect of protein topology and of static, energetic effects due to, e.g. pinning (polar water-binding) sites, and dynamic effects due to protein motion. Both of these slow down translation, but rotational retardation seems to come only from energetic (electrostatic) effects: when these are switched off, the water molecules actually reorient faster than in bulk. Translational motion happens by water molecules jumping between sites previously occupied by other waters, but also to sites previously occupied by protein groups – hence the involvement of protein motions. Without this protein motion, the options for water hopping are smaller, and so the diffusional retardation is even greater.

What about collective effects due to the hydrogen-boded network? These do seem to exist, and indeed to be strong – the effect of electrostatic pinning sites can percolate throughout the entire surface layer. That fits with the notion of percolation-dependent hydration dynamics discussed by Oleinikova et al. (Phys. Rev. Lett. 95, 247802; 2005).

This is a very nice paper that helps to prise apart the many factors operating simultaneously.

Water in tight places

A clutch of studies this week on water in confined geometries. Alan Soper and his coworkers at ISIS have used neutron scattering to look at the structure of water within the hydropholic pores of Vycor glass, the walls of which are lined with surface OH groups [J. Phys. Chem. B 111, 5610 (2007)]. For pores 40 angstroms wide – several dozen molecular diameters – they find significant disruption of the bulk liquid structure. The average number of hydrogen bonds per molecule is decreased from about 3.6 to about 2.2, and they see structural changes extending at least two layers into the liquid owing to the orientational effects of hydrogen-bonding to surface OH. It’s perhaps a little surprising that the effect is so big, and certainly raises questions about how bulk-like cell water is (the average distance between macromolecules in the cytoplasm is just 1-2 nm).

Sow-Hsin Chen at MIT and his colleagues have continued to study deeply supercooled water. Confined in mesoporous silica with pore diameters of 15 angstroms, it can be supercooled to at least 160 K. Deeply supercooled water is predicted to have a density minimum around 70 K below the well-known density maximum at 277 K – an echo of the ice-like low-density liquid phase predicted under pressure, and of the low-density amorphous ice that has been well established already. Using SANS, Chen and colleagues now see this density minimum at around 210 K for D2O [PNAS early edition, www.pnas.org/cgi/doi/10.1073/pnas.0701352104].

In a preprint [arxiv:0705.2348], Simone Melchionna use MD simulations to look at water in hydrophobic pores. They find spontaneous cavity formation for pore diameters of around 2 nm, which they relate to previous predictions and reports of density depletion and drying at hydrophobic surfaces, particularly the Lum-Chandler-Weeks model. In narrower pores (around 1.5 nm or so), this cavitation can result in complete (but intermittent) emptying, as has been proposed in protein ion or water channels as a gating mechanism. (In those cases, the precise nature of the residues in the pore neck seems to be rather crucial.)

Finally, something with a real biomolecule in it. Gerhard Hummer and colleagues have conducted MD simulations of water within the nonpolar cavities of tetrabrachion, a protein of the hyperthermophilic archaebacterium Staphylothermus marinus [JACS asap, doi:10.1021/ja070456h]. This contains several large hydrophobic cavities linked by a central channel, and the crystal structure shows that all contain water. The researchers find that the largest cavity contains 7-9 water molecules both at room temperature and at 365 K, the organism’s optimal growth temperature. But it empties at a slightly higher temperature, around 384 K. The second-largest cavity is filled with five waters at room temperature, as the X-ray structure confirms, but this breaks up at 365 K. Thus, both cavities are close to emptying at 365 K, and Hummer et al. suggest that this emptying might create sockets into which the proteases, which bind to it in its functional state, can plug.

More Hofmeister head-scratching

A paper by Dér et al. [J. Phys. Chem. B, doi:10.1021/jp066206p] makes the bold claim of providing a general microscopic interpretation of Hofmeister effects – the ion-specific salting-in or salting-out of proteins. I’m not sure that it succeeds. The basic idea is that the ions induce changes to the protein-water interfacial tension: so-called kosmotropes make the interface more hydrophobic, and chaotropes make it more hydrophilic. I can’t help feeling that the paper is hampered from the outset by an insistence on retaining the chaotrope/kosmotrope terminology, which was coined to suggest that the respective ions ‘break’ or ‘make’ ‘water structure’. There’s no evidence that ions do either, at least in terms of exerting any global influence on the hydrogen-bonded network. Indeed, Dér et al. acknowledge that spectroscopic studies [Omta et al., Science 301, 347; 2003] show no change in hydrogen-bonding on addition of ions beyond their first hydration shell. As far as I can make out, they seem to say that preferential segregation of ions at or away from the interface means that localized effects on H-bonding can be specifically felt there. But it’s not clear to me what they are thinking of in alluding to this surface segregation of ions – there’s no reference, for example, to the studies of that (at the air-water interface) by Jungwirth, Saykally and others. Nor do the authors seem to take into account how this picture applies to hydrophobic surfaces (Bruce Berne has studied this, and found ion-specific segregation). In any event, what results seems rather unsatisfactory, since we are then left with the confusing chao/kosmotrope terminology but a hint that in fact all the action is taking place at the interface (which is probably the case) – and an attempt to explain that action in terms of macroscopic interfacial tensions (which are not known anyway for protein-water interfaces). I can’t help thinking that it remains more useful to think more explicitly about how ions might modify the nature of the microscopic protein-water interface, and how this picture changes when two surfaces come together (so that, say, adsorbed ions are excluded).

Greg Voth’s group has just published a very nice review on proton transport in aqueous systems, including all the work on ‘water wires’ in aquaporin, M2, cytochromes and other proteins [J. Phys. Chem. B 111, 4300-4314; 2007].

Sylvia McLain and colleagues have conducted an extensive neutron-scattering study of the structure of proline solutions [J. Phys. Chem. B 111, 45568-4580; 2007]. Proline acts as an osmolyte or osmoprotectant, apparently protecting proteins against denaturation under water stress. It has been suggested that this is somehow due to proline clustering in solution, but McLain et al. find no strong evidence of that. Indeed, proline seems barely to perturb water’s hydrogen-bonded network at all, while remaining sufficiently hydrated to attain good solubility. They speculate that, despite the only weak tendency of prolines to cluster, they might form a protective sheath around proteins, chaperoning them to prevent denaturation.
Update: McLain and colleagues have just published a comparison of these experimental results with computer simulations (which show reasonable agreement): J. Phys. Chem. B 111, 8210; 2007.

There's a whole lot of simulating going on

Another biological ‘water wire’ is reported by Guillaume Lamoureux at U. Penn. and colleagues [Biophys. J. doi:10.1529/biophysj.106.102756]. They have simulated that ammonium transporter AmtB of E. coli, which has a hydrophobic channel that is thought to pass ammonia but to exclude water and charged species. The simulations, however, show that water can get inside and form a three-molecule chain. How this affects the permeation of ammonia (and exclusion of ammonium) isn’t yet clear, but it appears that the mechanisms proposed so far may run into problems.

Ulf Ryde at Lund and colleagues have simulated water in the active-site cavities of four human cytochromes: P450, 2A6, 2C8 and 3A4 [Rydberg et al., J. Phys. Chem. B, doi:10.1021/jp070390c]. In contrast to the crystal structures, they find that all the cavities are filled with water. That in 2A6 is small and contains only two waters, but the others have big cavities with around 40-60 water molecules – a volume of 1500-2100 angstroms. In these big cavities, water is rapidly exchanged with the environment through three to six channels. Those in 2A6 remain bound there, although quite mobile.

L-alanine acts rather like a surfactant in a water droplet, according to the density-functional MD simulations of Ivan Degtyarenko at Helsinki University of Technology and colleagues [J. Phys. Chem. B 111, 4227; 2007]. The amino acid moves to the droplet surface with its methyl group exposed, and the primary hydration shell of (on average seven) ordered and rather rigid water molecules forms around the carboxylate and ammonium groups.

More on the hydration of urea comes from Hinonori Kokubo and Montgomery Pettitt [J. Phys. Chem. B doi:10.1021/jp067659x]. They suggest that, to put it crudely, urea passes almost unnoticed in water – certainly, there’s no evidence of it acting as a structure-breaker. Another nail in the coffin for this hoary old idea.

Acid on top

[Victoria Buch and coworkers have just published a paper in PNAS on the acidity of the water surface, which is likely to have biochemical consequences. Here is the pre-edited article that I’ve written on the result for the May issue of Chemistry World.]

Pure water has an acid skin. This striking notion has been confirmed by calculations and tests by an international team of scientists. Victoria Buch of the Hebrew University of Jerusalem in Israel and her coworkers have found that the pH of pure water falls from the perfectly neutral value of 7 within the liquid to just 4.8 or less – about as acidic as beer – where water meets air at the surface [V. Buch et al. PNAS online, doi:10.1073/pnas.0611285104].

The finding could be significant for a number of disciplines. In the atmosphere, many important chemical reactions between trace gases take place at the surface of water droplets in clouds. “pH is an essential factor for many of these reactions”, says Pavel Jungwirth, one of Buch’s collaborators at the Academy of Sciences of the Czech Republic in Prague. “The low pH could also affect the rates of carbon dioxide absorption at the ocean surface”, adds water specialist Richard Saykally of the University of California at Berkeley.

And in molecular biology the effect might be reproduced where water comes into contact with water-repelling (hydrophobic) parts of proteins, changing the acid-base chemistry that goes on at protein surfaces. “The effects of interface acidification on protein and membrane structure could be huge”, says theoretical chemist Gregory Voth at the University of Utah.

“I think that this is a very important paper”, says Saykally, who has previously found indirect evidence of the surface acidification. While the prediction was well-known, he says, “this is the first study to actually predict the pH of the liquid water surface.”

It has become clear in the past several years that the water surface may look very different from the bulk liquid. Saykally and Jungwirth have shown that dissolved ions can become either depleted or concentrated at the surface [e.g. P. B. Petersen & R. J. Saykally, Annu. Rev. Phys. Chem. 57, 333; 2006]. But the accumulation of protonated water molecules (H3O+ or hydronium, the basic component of acidity in water) at the air-water interface happens for unique reasons.

In 2004 Voth and his coworkers used quantum-mechanical computer simulations to show that hydronium prefers to sit at the water surface rather than deep inside the liquid [M. K. Petersen et al. J. Phys. Chem. B 108, 14804; 2004]. Whereas H2O molecules typically form four hydrogen bonds to their neighbours – two via the hydrogen atoms, and two via the electron lone pairs on oxygen – H3O+ can only form three. The three hydrogens can bind to water molecules, but the oxygen atom, where most of the positive charge resides, can’t any longer act as a good ‘acceptor’ for hydrogen bonds.

This makes it energetically unfavourable for the oxygen side of hydronium to be in water at all. So, Voth’s team said, hydronium acts somewhat like an amphiphile – a molecule with a water-soluble part and a hydrophobic part, like a soap molecule. The ions gather at the air-water interface with the hydrogens pointing downwards to make hydrogen bonds and the oxygen pointing up out of the liquid [S. S. Iyengar et al. Int. J. Mass Spectr. 241, 197; 2005].

In 2005 Saykally and his colleague Poul Petersen used a spectroscopic technique to adduce indirect evidence for this surface excess of acidic hydronium [P. B. Petersen & R. J. Saykally, J. Phys. Chem. B 109, 7976; 2005]. Buch and colleagues now have fresh evidence, albeit still somewhat indirect: they show that deuterated water D2O can swap hydrogens with H2O about 20 times faster at the surface of ice nanocrystals (which have a liquid-like surface) than deeper inside. This switch depends on the presence of hydronium ions.

The researchers also have new simulation data for the surface acidification, which enables them to estimate the pH increase. But Voth cautions against placing too much weight on this number as yet, because of the simplifications in the model. For example, “they have no possibility of having the hydronium and hydroxide recombine into a water molecule”, he says.

A key question now is whether the experimental evidence could be made more direct – “ice crystals are not that relevant to water.” says Voth. But this will be a challenge: “it is very hard to measure the top surface layer of a volatile liquid like water”, says Jungwirth. Saykally agrees with that. “We are thinking hard about how to do this,” he says, “but it’s not easy.”

Smooth folding

A paper in the forthcoming issue of PNAS (104, 6206; doi/10.1073/pnas.0605859104) by Peter Tieleman’s group at the University of Calgary looks at how to reconcile the energy-funnel picture of protein folding with the fact that there are likely to be significant enthalpic barriers to the folding process. They use simulations of the association of two polyalanine and two polyleucine alpha-helices to figure out whether enthalpic barriers exist (they do), and why. It seems they arise in this case from the fact that, as the chains approach, they must become desolvated before there is a compensating enthalpic gain from strong helix-helix interaction. (Interestingly, the researchers see no dewetting transition as the helices approach, of the sort predicted by Lum, Chandler and Weeks (J. Phys. Chem. B 103, 4570-4577; 1999), but only ‘steric dewetting’ when there is simply no longer space to fit in a layer of water. This contrasts with the simulations of hydrophobic plates by Bruce Berne mentioned in my previous post (JACS asap doi:10.1021/ja068305m), where dewetting does feature. It is possible that this might be rather sensitive to the precise geometry, size and hydrophobicity of the two surfaces.) The enthalpic energy barrier is, however, largely compensated by the gain in solvent entropy on desolvation, leading to a free-energy barrier that is very small (for poly-A) or non-existent (for poly-L). Thus, the idea of a relatively smooth free-energy funnel is recovered.

I’ve now had a better look at the Berne paper. It suggests that Hofmeister effects are indeed complicated, and not best explained by a simplistic structure-making/breaking model. In the case of the association of nanoscale hydrophobic surfaces, the effect of ions depends on whether or not they accumulate preferentially at the surfaces. High-charge-density ions induce salting-out (reducing the solubility of hydrophobes) via an entropic effect due to preferential exclusion of ions from the interfaces. Medium-charge-density ions induce salting in because of a different entropic effect, due to strong hydration of the ions and a consequent reduction in solvent entropy when the ions, preferentially adsorbed at the surfaces, are expelled as the surfaces associate (I think I’ve got that right). But low-charge-density ions cause salting in enthalpically, since they bind to the surfaces and lower the surface tension of the plate-water interface, thus lowering the enthalpy of association. As if this isn’t complicated enough, Berne and colleagues say that something quite different applied for electrolytes and small hydrophobic particles (see Zangi & Berne, J. Phys. Chem. B 110, 22736-22741; 2006). Hmm.

Not strictly related to biochemistry, but Rudy Marcus and Yousung Jung have just published a proposed explanation for why there is a rate acceleration of certain organic reactions, particularly those associated with ‘click chemistry’, at the interface of water and an organic phase: see JACS asap doi:10.1021/ja068120f.

Why cell fluid is lumpy

How homogeneous is the cytoplasm? There is increasing evidence that proteins in concentrated solution form relatively long-lived clusters. Wilson Poon and his colleagues showed in 2004 that this effect, previously rather anecdotal, is real and general, applying to colloidal particles as well as proteins (Stradner et al., Nature 432, 492; 2004). They showed using small-angle X-ray and neutron scattering that lysozyme forms clusters of about 3-10 molecules at volume fractions of between 0.05 and 0.2, which they say are equilibrium structures resulting from the interplay of short-ranged attractive (van de Waals) and electrostatic repulsive forces. Now Peter Vekilov at Houston and his coworkers broadly support that notion by looking at concentrated solution of bacterial lumazine synthase using light scattering (Gliko et al., J. Phys. Chem. B 111, 3106; 2007). They see clusters with lifetimes of around 10 s and a mean radius of about 350 nm (individual molecules are bout 15.6 nm in diameter). Changing the protein concentration changes the cluster concentration (which can reach a volume fraction of 0.001), but not the cluster size. But Velikov et al. say that cluster formation and size is dominated by kinetics, not thermodynamics: these clusters are metastable with respect to both protein crystals (i.e. in supersaturated solution) and to well-dispersed solution.

All of this recalls the flurry of interest in clustering excited by the work of Geckeler and Samal on C60 (Chem. Comm. 2001, 2224), which was rather breathlessly touted as a possible mechanism for homeopathy. That work was truly odd, as it seemed to suggest that the cluster size increased with increasing dilution. I’m not aware that the result has been reproduced. Needless to say, the homeopathy connection makes no sense (at best, you get a few ‘active’ bottles and the vast majority containing just water); but at the very least, there’s no reason to regard clustering per se as perplexing or odd.

The question of a vapour gap at the interface of water and a hydrophobic surface, and its relation to the long-ranged hydrophobic attraction, seems to be resolving itself. It seems now that a very thin depletion layer exists. But the role of dissolved gases in forming nanobubbles remains to be fully resolved (see ‘Why does water do that’ below). They have been seen by various methods, but with the proviso that they could possibly be an artefact of the probe technique, and that they haven’t been shown definitively to be gaseous anyway. Also, small nanobubbles should have a small radius of curvature and thus a large Laplace pressure, promoting their dissolution. William Ducker and colleagues at the University of Melbourne have now shown that flat gas bubbles, about 5-80 nm thick and 4 microns across, can exist at such a hydrophobic interface for over an hour (Phys. Rev. Lett. 98, 136101; 2007). This size means that the internal pressure is barely above atmospheric. But the bubbles form only when a particular protocol is followed for introducing the gas layer (carbon dioxide): in other words, “the presence of the gas phase depends on the previous history of the interface.”

Ivan Brovchenko and colleagues recently linked the low-hydration polymorphic transitions of B-DNA to the presence (or not) of a fully connected (percolating) network of water molecules in the hydration sphere (Brovchenko et al., Phys. Rev. Lett. 97, 137801; 2006). They’ve now extended that work by looking at the percolation transition for both B- and A-DNA (Brovchenko et al., J. Phys. Chem. B 111, 3258; 2007). Although the percolation thresholds (that is, the surface coverage of water on the DNA molecules) are virtually identical, the mechanisms are quite different in each case: the threshold corresponds to the appearance of a spanning water network in the major groove of B-DNA, but the minor groove of A-DNA. It isn’t clear, then, whether the near-coincidence of the two thresholds is indeed just coincidence or has some deeper physical cause. In any event, there are also insights here into how ions can alter the hydrogen-bonding patterns and thus shift the thresholds.

Water seems to play an important role in electron transfer between some protein redox centres – this was discussed nicely by Gray & Winkler recently (PNAS 102, 3534; 2005). They enumerated the various ways, direct and indirect, that water might facilitate electron hopping. Agostino Migliore and colleagues in Modena have now used ab initio calculations for the copper active sites of azurin to figure out how water-mediated pathways are functioning in this case (Migliore et al., J. Phys. Chem. B, advance online publication, doi:10.1021/jp068773i). But it’s a complicated picture that emerges, in which no one physical pathway seems to be responsible for what is observed. Sorry, but it’s hard to put this one into more of a nutshell than that.

Finally, and perhaps even more cryptically, I want to flag up a paper by Bruce Berne and colleagues in JACS ASAP (doi:10.1021/ja068305m) on the effect of ions on the hydrophobic interaction between two plates. This complements Berne’s recent study of much the same thing for hydrophobic particles (J. Phys. Chem. B 110, 22736; 2006). The phenomenon is of course intimately related to salting-in/out and Hofmeister effects, and as such, contains a lot of important information on the effect of electrolytes on protein aggregation and folding. So there’s a lot of good stuff here to digest, and I’m not going to manage that in a hurry without risking indigestion (or more probably, misapprehension). Worth spending time on.

How thick-skinned is hydration?

How much water do you need to fully solvate a protein? There are many studies of protein behaviour at low water coverage, going back to the suggestion by Rupley and Careri (Adv. Protein Chem. 41, 37-172; 1991) that proteins seem to require about 0.4 g of water per gram of protein to achieve their normal functionality. Roland Winter and coworkers have investigated the notion of a percolation transition in water coverage that brings the protein dynamics to life (Oleinikova et al., J. Phys. Chem. B 109, 1988-1998; 2005; Smolin et al., J Phys. Chem. B 109, 10995-11005; 2005). Mehdi Bagheri Hamaneh and Matthias Buck at Case Western have looked at the question in a rather different light: how much water do you need to put around a protein in order to be able to simulate it realistically? They find (Biophys. J. 92, L49; 2007) that you don't need to fill up your simulation box with explicit water – a shell just two or three layers thick (using the CHARMM22/CAMP potential function) will do the job well enough. That's computationally cheap, and I suppose implies that there's not really much excuse for failing to model hydration explicitly. It also implies that the celebrated cooperativity of water dynamics in the hydration shell does not appear to extend very far – at least, perhaps one should say, for the case of lysozyme considered here.

More on the 'glass transition' at around 220 K, seemingly shown now (Chen et al., PNAS 103, 901; 2006) to be a fragile-to-strong crossover. This applies also to DNA (Chen et al., J. Chem. Phys. 125, 171103; 2006), and now Sow-Hsin Chen at MIT and coworkers have found similar behaviour for RNA, again at 220 K (http://xxx.arxiv.org/abs/physics/0703166). So this seems to be pretty universal behaviour for biomacromolecules, reinforcing the idea that the change in dynamics is imposed by the hydration water.

Masahiro Kinoshita has sent me a couple of his papers from Chem. Phys. Lett. (see them here and here) which explore his idea that "the major driving force in protein folding is a gain in water entropy". In a nutshell, they say that "a protein is designed to fold into the structure that maximizes the entropy of water under the requirement that sufficiently many intramolecular hydrogen bonds be formed to compensate the dehydration penalty." In other words, as I understand it, the enthalpies balance and what's left (governing stability) is the water entropy change. That's intriguing; I'm still struggling to see how this ties up with Jack Dunitz's suggestion (Science 264, 670; 1994; Chem. Biol. 2, 709-712; 1995) that transferring a water molecule from an ordered binding site where it is bound by an ‘average’ hydrogen bond to the bulk involves an overall free-energy change that is close to zero, and with ideas about the importance of interactions at specific residues – the 'hotspots' discussed in the last post, and Ariel Fernandez's notion of dehydrons, for instance. One day, perhaps, it will all make sense to me.

Why are hot spots hot?

In protein-protein interactions, some residues do nearly all the work. These 'hot spots' are apparently responsible for most of the binding energy, something that becomes apparent if they are mutated to alanine. It has been suggested that hot spots rely on being dry – sheltered from water by a surrounding ring of protective residues, called the O ring. But that idea hasn't been well tested. Now Maria Ramos and coworkers at the University of Porto in Portugal have studied it using MD simulations (see paper here). They find that indeed hot (and 'warm') spots in the interaction of an immunoglobulin and a lysozyme do seem to be relatively inaccessible to water. Moreover, the water molecules that do get past the O ring have relatively short residence times, much the same as those of bulk water – they don't particularly want to be there.

Greg Voth and his coworkers have a paper in JACS looking at proton transport in cytochrome c oxidase. It has been suggested that a proton is held in a 'trap' in the bovine form of this enzyme before being transported once a residue elsewhere is deprotonated. Voth and colleagues show that this process happens for a bacterial cytochrome c oxidase (Rhodobacter sphaeroides) too, and that it depends on a hydrogen-bonded network of about five water molecules that is somewhat comparable to the proton-release and transport complex found in bacteriorhodopsin (see Garczarek & Gerwert, Nature 439, 109; 2006).

Water from first principles?

A paper in Science this week claims to present a new, improved effective pair potential for water calculated from first principles, which works well both for water dimers and for the bulk liquid. I was a little puzzled by this, since there has been plenty of previous work in this area and it wasn’t immediately clear what the new trick was here that had enabled the claimed improvements. Having been asked to write about the work for Chemistry World, I contacted two experts in this field. Both were strongly critical of the paper.

First, the general context, which is nicely explained by one of my advisers:

“Modeling intermolecular interactions in water in terms of simple pair potentials is difficult because these potentials miss the cooperative effects of the H bonds. These are important: it is because of these effects that the “dipole” moment of a molecule in condensed phase is greatly enhanced compared to that of a molecule in gas phase (here I put dipole in quotes because this quantity cannot be unambiguously defined in condensed phase). Compared to simple liquids water has a much more open structure dictated by the underlying H-bond network.

Popular empirical potentials for water try to model the basic physics underlying H bonds interactions, which are largely of electrostatic origin in terms of interactions between point charges spatially located in a way that mimics, albeit very approximately, the charge distribution in the water molecule. These potentials are tuned to reproduce a number of properties of the liquid but have limited transferability to different environments as they miss the cooperative effects mentioned above.”

Existing classical potentials for water describe well the structure of the liquid (even very well I would say!). The difficulty with this approach, apart from the issue of transferability, is that we also need to describe the interaction between water and other molecules, material systems etc. in order to model solvation, interfacial water, confined water etc. In addition, there are properties such as the dielectric properties that depend on the actual electronic structure of the molecules in a given environment. Furthermore water molecules can dissociate. A good potential for the intermolecular interaction does not tell us everything.”

But in terms of the Science paper by Bukowski et al., he says:

“I am not too impressed by the contribution of Bukowski et al. They show that an extremely good potential for the dimer (a pair potential) is not sufficient for a good description of the liquid, as one could have expected. Including non-additive many body effects they obtain a more decent agreement with the experimental liquid structure. Interestingly, and to some extent unexpectedly (at least to me), simple polarization effects seem to be doing most of the job. However, in the end they have just another rigid water potential which, judging from their pair correlation function, appears to be almost as good as the best existing rigid empirical potentials for water. Of course conceptually it is not the same thing: the potential of Bukowski et al. is not tuned to reproduce experiment but is derived from accurate quantum mechanical calculations on the dimer (and in the most accurate case also on the trimer). This is an important achievement but difficult to generalize to a wide range of possible contexts, including e.g. flexible molecules, solvation effects, hydrophobic and hydrophilic conditions, confinement, systems other than water, etc. Keeping the attention on the liquid structure, I do not think that this potential gives us a better insight than what we already know on the structure of the H bond network in water.

I do not think that the work of Bukowski et al. goes anywhere beyond ab-initio molecular dynamics on the flight. [This is the kind of approach used by Dominik Marx, Roberto Car and others.] The latter models water in terms of nuclei and electrons, the former in terms of rigid intermolecular potentials. The latter produces flexible molecules by construction and describes in detail the interplay between electronic structure and nuclear dynamics. As such ab-initio molecular dynamics on the flight is applicable to the most general range of situations, including for instance proton transfer effects (the Grotthus mechanism) that are not allowed by any rigid classical intermolecular potential (whether ab-initio or empirical!). The main limitation of ab-initio molecular dynamics on the flight, apart from numerical cost, is due to the limited accuracy of existing approximations of density functional theory, but as these improve or if highly efficient and more accurate electronic structure methods are established, immediately this progress could be transferred to the accuracy of ab-initio molecular dynamics on the flight. It is not so with the method of Bukowski et al. which cannot go beyond the accuracy provided by a simple polarization approximation of the non additive many-body effects. The only system that Bukowski et al. describe better than ab-initio MD on the flight is the potential of the water dimer, for which they rely on the most accurate available quantum chemical methods. Some of this accuracy is lost when they go to condensed phase. Judging from their current results, they do not have a better description of liquid water than that provided by existing empirical potential. Overall, the insight on a number of properties, structural, electronic etc. provided by ab-initio MD (on the flight) is vastly superior.”

That squares with my second adviser, who says:

“I'm very disappointed such a paper is appearing in Science. First, it gives an incredibly misleading account of the existing ab-initio literature on water, almost entirely based on DFT, and it omits a very recent quantum chemistry work (by S. Xantheas, JCP). Second, it "sells" a quantum chemistry based ab-initio potential as giving excellent agreement with experiment, when the agreement the authors find is not much better than what is already out there, in my opinion, using DFT based methods. This Science work just has different kind of disagreement with experiments, w.r.t. those found with DFT, but overall it is not much better (actually it is a little worse); most importantly: this work does NOT solve any of the open problems out there on the structure and/or properties of water.

Let me elaborate a bit on these points.

1) Account of the existing ab-initio literature. In the last ~ 15 years, most of the ab-initio simulations (no fit to or input from experiment) of water have been performed using Density Functional Theory (DFT) based methods, and they have been carried out mainly with two functionals (so call gradient corrected energy functionals, BLYP and PBE). Both of these functionals have been believed to give good agreement with experiment for several years (specifically until ~ 2004). However the good agreement with experiment did NOT came from a good performance of the theory but was somehow fortuitous, due to numerical inaccuracies in the solution of the Kohn-Sham equations (DFT equations) for the electrons. In 2004, Schwegler et al. (JCP 2004) and Grossman et al. (JCP 2004) pointed out these inaccuracies and showed that water correlation functions (g(r)--also discussed in the Science report) are over-structured and diffusion too slow (w.r.t. to experiment), when the numerics is done right and numerical inaccuracies are removed. The fortuitous agreement with experiment there had nothing to do with a fortuitous choice of functionals, as stated by the present authors. It had to do with numerics adopted in integration techniques. Those results were later confirmed by Sit and Marzari (JCP 2005), Serra and Artacho (several papers appeared in JCP) and others (none of these papers are mentioned in the Science report). So the introduction of this Science paper carelessly dismisses an approach (ab-initio simulations based on DFT) that, although not in full quantitative agreement with experiment, can describe well the salient, qualitative features of liquid water. On top of this, at the end of the paper the reader realizes that such a dismissed approach gives an agreement with experiment which is similar to the one found by the authors (Actually, I'd like to claim that DFT gives a more consistent agreement with experiment than the one presented in the Science paper--see below).

I also note that in the introduction of this paper dispersion forces in water are declared "non negligible", with no reference and no discussion. There is no clear experimental and/or theoretical evidence to support this statement. It may well be so, but nobody knows right now.

2) Agreement with experiment found here, wrt to existing ab-initio simulations based on DFT. I believe that overall the g(r)s found here are not in better agreement with experiment than those described by DFT ( see original papers). Note that they get right the first peak of g(r) but not the second, implying that their angular distribution (if they had computed it!) is quite inaccurate. Their diffusion coefficient is in agreement with experiment, but admittedly (see their own statement), this agreement may be fortuitous as they neglected proton quantum effects of the monomer, and flexibility of the monomer. Most importantly: the coordination number they find is an overestimate of what is accepted in the field as a reasonable number extracted from experiment. They find 5.6 and they compared it with 4.8, when the accepted value in the field is more like 4.3/4.5. Whatever number you consider, their value is a big overestimate, giving a liquid over-coordinated with respect to experiment and with too many hydrogen bonds. Their computed internal energies are worst that those obtained with empirical potentials.”

Not encouraging, then. And in any event, in terms of understanding hydration, the messages were as follows:

Adviser 1: “I cannot predict what would be the result of applying the scheme of Bukowski et al. to studying hydrophobic hydration. In hydrophobic hydration there is not just water but also the hydrophobic substance (unless this is the vacuum). Since their potential should be more transferable than standard empirical potentials, it should describe better the water close to a hydrophobic solute. In principle, however, all the interactions need to be included and treated with comparable accuracy. In our study of solvated methane both the water and the methane molecules have deformable electronic clouds that play an important role in the outcome of the calculation. These effects are described to some extent in terms of polarization effects by Bukowski et al., but they should also construct a methane-methane and a methane-water potential and include polarization effects beyond pure water in order to tackle the solvation problem.”

Adviser 2: “Even if the authors found the best ab-initio potential for liquid water fitted to ab-initio gas phase data, if they wanted to describe solvation they would have to start all over again, as they would have to do the fitting to gas phase water containing the solvated molecule or they would have to add other pieces to the potential. This is why fitted potential (whatever they are fitted to: ab-initio data, experiment, etc.) will always have serious drawbacks. If you want to study systems that have not been fitted, containing other entities, well... here you go, you have to start again with your fit.”


So there it is. I don’t like dumping on a paper, but this one has come out in a very high-profile journal where it will get a lot of recognition. So I thought it is only right to put the record straight.

Dynamics fast and slow

How do motions of a peptide chain depend on those of water molecules in the hydration shell? It's clear enough that the solvent dynamics can span a wide range of timescales, depending on how the water molecules interact with the protein. But in unfolded and molten-globule states, it has been suggested that there are rapid fluctuations between various helical and beta-sheet-like states that are 'lubricated' by picosecond rearrangements of the hydrogen-bonding network in water. This lubrication, enabled by the rapidity of H-bond making and breaking, is presumed to enable protein folding. Neil Hunt and coworkers at the University of Strathclyde now say that they've seen such motions, with timescales of a few tenths of a picosecond, in the alpha-helix-to-random-coil transition of a homo-polypeptide (poly-L-lysine) using optical Kerr-effect spectroscopy. The paper is here.

Motions in the hydration shell that are one or two orders of magnitude slower have been studied by Dongping Zhong and colleagues at Ohio State, in a paper here. They're looking at apomyoglobin, and specifically at the water dynamics around a tryptophan group (Trp7), which serves as a convenient fluorescent chromophore. Through both experiment and simulation, they find slow relaxation on timescales of 5-87 ps. Previous studies of such slow dynamics have offered divergent interpretations. Ahmed Zewail and coworkers have suggested that these dynamics are due to the effect of the protein's potential field on the hydration water (J. Phys. Chem. B 107, 13218; 2003). Bertil Halle thinks that these water dynamics close to a protein aren't so different from those in the bulk (PNAS 102, 13867; 2005). Zhong and colleagues say that the slow relaxation is due to strongly coupled water-protein motions. If either the water or the protein is frozen in the simulations, the slow component disappears. I guess that supports the contention of Bizzarri and Cannistraro that the dynamics of the protein and solvent are so strongly coupled that they ‘should be conceived as a single entity' (J. Phys. Chem. B 106, 6617-6633; 2002).

Quantum hydrophobicity, and water in drug design

A paper in PNAS by Ned Wingreen at Princeton uses first-principles quantum molecular dynamics to look at the hydrophobic interaction of two hydrated methane molecules. I’d say this is primarily a methodological paper – that’s pretty much the angle Guilia Galli takes in her accompanying commentary – in that it aims to establish how well classical simulation approaches do in capturing the nature of the interaction. The answer is: not particularly. Classical force fields predict two free-energy minima, one at methane-methane contact and a second, rather shallow, at a small separation corresponding to one intervening solvent layer. But the relative stability of these minima is rather sensitive to the precise force-field parameters and can be reversed for some values. The quantum simulations reveal a deeper potential well at contact – this is always the stable configuration. But there is a succession of shallow minima at several other methane-methane separations, implying the existence of several relatively stable hydration cages. The authors talk about these configurations in terms of ‘clathrate-like cages’, but in fact it seems that they don’t have any well-defined hydrogen-bonding arrangements: two independent simulations at a specific methane-methane separation gave water structures that could not be superimposed. (There was, however, apparently some consistency in the ‘hydrogen-bonded rings’ in between the two methanes.) Wingreen and colleagues suggest that the shallow minima are the result of relatively well packed configurations for water in the hydration shells. But I don’t really know what this means. Normally, considerations of molecular packing in liquids are governed by the short-ranged repulsion between molecules. But ‘well packed’ is an ambiguous term for water, where optimal hydrogen bonding means that the waters sit rather further apart than equivalent spherical molecules would do. I’m assuming ‘well packed’ here refers to unbroken, unstrained H-bonding configurations…?
**********

“Until a few years ago it was common practice to ignore water molecules in protein binding sites”, say Jonathan Essex and colleagues at Southampton University in an interesting paper in JACS. But now, they point out, there is increasing interest in designing ligands that will displace particular water molecules in drug binding. Conceivably, this might make the ligands more selective and the binding energy more favourable.

But it’s not clear whether that will necessarily be so. Despite the entropic advantage of expelling bound water from a binding cleft, one can’t generalize about the consequent free energy change. Whether or not it is advantageous to incorporate a water molecule at the binding interface hinges on a delicate balance. Confining a water molecule clearly has an entropic penalty, but this might be repaid by the enthalpic gains of hydrogen-bond formation – an issue that must itself be weighed against the average number of hydrogen bonds that a bulk water molecule engages in. Jack Dunitz (Science 264, 670 (1994); Chem. Biol. 2, 709-712 (1995)) has estimated that transferring a water molecule from an ordered binding site where it is bound by an ‘average’ hydrogen bond to the bulk involves an overall free-energy change that is close to zero. So it is not obvious which way the scales will tip in any instance. John Ladbury and his coworkers (D. A. Renzoni, M. J. J. M. Zvelebil, T. Lundbäck & J. E. Ladbury, in J. E. Ladbury & P. R. Connelly (eds), Structure-Based Drug Design: Thermodynamics, Modeling and Strategy 161-180, Landes Bioscience (1997)) have thought about the implications for drug design.

The message is illustrated in the binding of various inhibitors of HIV-1 protease, one of the key targets in AIDS therapies. Crystal structures show that some of these, such as KNI-272, bind to the enzyme via a bridging water molecule. Other inhibitors, such as DMP450, have been designed specifically to exclude this water molecule while mimicking its hydrogen-bonding capacity, and have found to bind more strongly. Li and Lazaridis (JACS 125, 6636-6637 (2003)) have calculated that displacement of the bound water by DMP450 is in itself unfavourable relative to KNI-272, but that this cost is outweighed by the lower cost of desolvating DMP450 to form the bound complex. So the consequences of eliminating the water molecule are both highly specific and not obvious.

With all this in mind, Essex and colleagues have sought a way of classifying water molecules in protein binding sites according to how easily displaced they are by ligands. By studying the thermodynamics of six proteins complexed with a variety of ligands, they say that the water molecules in the binding sites seem to come in two classes: those that are readily displaced (by at least some ligands), and those that never are. The latter, unsurprisingly, turn out to be more tightly bound according to MC simulations. All the same, the authors say that “no linear correlation exists between the binding free energies of water molecules and the change in binding affinity of ligands displacing the water molecules.” Yet they conclude that if we can identify the ‘conserved’ water molecules – those that do not get displaced come what may – then these can be usefully used in the design of drug docking: in effect, they serve as ‘part of the protein’, available for hydrogen bonding to the ligand. This paper supplies some heuristics for deciding which water molecules are conserved or displacable.

Why does water do that?

Following on from the suggestion that the ‘glass transition’ of hydrated proteins at around 220 K is in fact a change in the dynamical state of the water of hydration, related to the second critical point of water in the supercooled, high-pressure state (see the previous post below), Gene Stanley and his coworkers have now released a preprint that aims to relate this behaviour to the molecular structure of the liquid. To recap, the idea is that the ‘transition’ corresponds to the crossing of the so-called Widom line, the locus of the maximum in the correlation length that extends like a ‘ghost’ of the liquid-liquid phase transition beyond the critical point at which this transition vanishes. (At the critical point itself, this correlation length diverges.) What does this mean for the nature of the hydrogen-bonded network? Well, it’s subtle. In thermodynamic terms, the crossover corresponds to a change from non-Arrhenius dynamics at high temperature (the activation energy for water diffusion depends on temperature) to Arrhenius dynamics at low temperature (temperature-independent activation energy). The simulations by Stanley and colleagues now suggest that this crossover shows up in terms of the tendency to form non-bifurcated hydrogen bonds. Bifurcated H-bonds are fairly common in the liquid state at ambient conditions, and seem to be the defects that make diffusion facile. The Widom line corresponds to the point at which the derivative of the probability of forming non-bifurcated bonds with respect to temperature is maximal. Hmm, not an easy quantity to visualize. But it does mean that above the Widom line, supercooled water has fewer non-bifurcated hydrogen bonds, and so is less ‘tetrahedral’ and denser (like the high-density liquid phase), than below this line. That’s intuitive enough – crossing the Widom line as temperature decreases corresponds to the supercooled liquid becoming less dense, more structured and changing from a high-density-liquid-like to a low-density-liquid-like state: a ghost of the liquid-liquid phase transition itself, but with no abrupt change of thermodynamic variables.

Lawrence Pratt at Los Alamos and his coworkers have posted a preprint entitled ‘What is special about water as a matrix of life?’. This, as I recall, is basically the paper that Lawrence presented at the Varenna meeting, convened to discuss that very question in early 2005. The paper aims to address the title question in general, and hydrophobic effects in particular, by moving away from an emphasis on structure (for example, the classic Kauzmann model of entropically driven hydrophobic attraction due to the release of ‘structured water’ in the space between hydrophobes) and focusing instead on what the authors call the ‘engineering characteristics of the liquid’, such as its equation of state. The key message seems to be that, as a solvent for life, water represents a safe bet: the liquid state exists over a wide temperature range (compared with other simple small-molecule solvents), and within that range there is rather little variation in thermodynamic variables: response functions such as the compressibility and thermal expansion coefficient, as well as the nature of the solvophobic effect, vary little. It seems to me that this raises several questions (which are not objections), such as: does life require a wide temperature range, or does it just fill up whatever niches are available? (Thermophiles seem very ancient; would life have happily persisted if all Earth’s water was warm?) Does the ‘structured’ character of water play any necessary role in this scheme? (It does seem to be biologically important that water forms directional H-bonds.) And what does underlie the hydrophobic attraction, and how general is it?

Speaking of which, more evidence for the role of nanobubbles in the long-range hydrophobic attraction is provided in a study of nanoparticle adsorption onto a quartz microbalance in gassed and degassed water by Sangmin Jeon and colleagues in Korea (Langmuir 23, 1623-1625 (2007)).