To Hofmeister and beyond: an improved understanding of protein solubility and stability

Proteins are used increasingly widely, in pharmaceutical and diagnostic formulations, cosmetics, industrial processes, and detergents. A major problem is that proteins have limited solubility and/or stability in liquid formulations, leading over time to loss of quality, reduced benefit, wasted cost (especially as the protein is usually the most expensive part), and a lack of reproducibility and predictability. Similar problems abound in other spheres: for example many proteins give poor analytical spectra, cloudiness or haze in beverages, and cannot be crystallised, because of difficulties in keeping them in solution. There is thus a major need to improve stability without losing solubility or vice versa. The main approach used is to add high concentrations of ions and other molecules to the protein solution (known generally as excipients). Currently the selection of excipients is usually tackled on an ad hoc basis, using high-throughput testing of different solution conditions, and there is very little theoretical rationale that could be used as a basis for a more rational, cheaper and quicker route to improved performance. This proposal aims to provide a more rational basis for understanding and improving stability and solubility.

We recently proposed a model that explains how salts added to a protein solution affect its solubility and stability, by competing with the protein for water molecules. Different salts compete better or worse than the protein, and therefore have different effects. The effects have been known for over 100 years and are generally called the Hofmeister effect, but there remains considerable disagreement over the explanation. This matters, because once we understand the physical basis for the Hofmeister effect, we can exploit it to develop better solution conditions. Our model was based largely on measurements using a spectroscopic technique known as NMR.

The model differs in important ways from the current standard models. We will therefore start by using NMR to compare our model to the most popular of these, to eliminate the most significant likely objection to our model and clear the ground. Our model emphasises the role of water molecules in mediating changes in solubility and stability: in particular, it predicts that ions that stabilise proteins necessarily make them less soluble, and vice versa. If this is strictly true, then it places severe limits on what we can do to (for example) improve stability without compromising solubility. We shall therefore explore ways of getting round the problem. First, we shall investigate whether mixtures of ions behave simply in an additive way. To the extent that they do not, we might be able to exploit a window of opportunity. A particularly interesting pair of ions is a mixture of the two amino acids arginine and glutamate, which seem to work in a different way and are apparently 'special'. We shall investigate how they work, whether they are in fact different, and whether combining them with more typical Hofmeister ions helps. We will also investigate our suspicion that increasing the size of a protein makes the effect of added ions weaker: if true, this is useful information, since it guides the way we would go about stabilising different proteins.

There are many organisms that can grow in high salt concentrations. Normally, high salt reduces either solubility or stability or both: so how do these organisms manage to grow successfully? The answer is that they produce high concentrations of specific small molecules to balance the high external ionic strength. We shall investigate how these molecules work, and whether they behave in the same way as 'normal' ions. If they work in different ways, then these can be exploited.

Finally we will study whether external pressure can be used as an alternative way to stabilise proteins, and thus whether pressure would be a useful variable. The outcome will be a rational toolbox to guide solution conditions.

This proposal builds on our recently submitted model describing how Hofmeister anions act to stabilise or solubilise proteins, which to a large extent goes back to the classic explanation, that they make or break the water structure, and thereby affect the thermodynamics of protein solvation. It aims to explore how Hofmeister ions work, and to examine possible non-Hofmeister effects, in order to develop a more rational and predictive basis for increasing either the solubility or stability of proteins without worsening the other. The methodology builds on our previous results, using NMR, DSC and DLS, mainly on the protein barnase, a simple and well-characterised model system. First, we investigate whether low charge density ions bind and stabilise hydrophobic surfaces, since if they do (which we doubt, although competing theories suggest this) then we need to think much more about the thermodynamics of the unfolded protein. Then we study whether binary mixtures of anions are simply additive - how 'ideal' is Hofmeister? A key part of the study is an investigation of equimolar arginine/glutamate additives, which for some proteins cause a dramatic increase in stability. Is this 'special', working by a non-Hofmeister mechanism? We will use NMR and DSC to see if it fits the Hofmeister pattern - and if not, we shall find out what it does, and investigate related pairs to see if it is unique or part of a predictable pattern. If it does work by a non-Hofmeister mechanism, does simultaneous addition of Hofmeister anions produce predictable additive effects? If so, this would allow us to alter stability without significantly affecting solubility, or vice versa.

We also study other ways of stabilising/solubilising proteins, namely protein size/oligomers, osmolytes such as betaine, and increased pressure. The combined results will allow us to predict what types of additives will produce the desired improvements in stability or solubility, which will be of wide applicability.

The project is aimed at improving our understanding of protein solubility and stability, with a view to identifying general routes to better solubility and stability in general, tailored where necessary to particular proteins or types of proteins. Therefore as set out in the section Academic Beneficiaries, we expect this project to have a large impact, both in the scientific community generally and more specifically within industry.

It is worth commenting specifically on the impact to academics. Most academics who work with proteins will have encountered problems with solubility or stability, but most just put up with it, or else try whatever local magic they may have come across, and then give up, without even (for example) carrying out a literature review of techniques for improving protein stability, on the assumption that if there were useful ideas out there, then someone would have already tried them, and if they were useful then they would be widely known. So our challenge here (assuming that we are successful) is in getting the message out. Naturally we shall aim to publish the research in high impact journals. We shall also target conferences where our message is likely to be heard, eg US conferences, particularly of the Protein Society, but also meetings with high industrial input such as the BioProcessUK and MIBio meetings in the UK, both of which we have spoken at in the past. We have requested a budget for conference attendance, for this purpose. The University of Sheffield has a very active Communications team, who are good at getting notice for press releases and web announcements. MPW is the web manager for the Molecular Biology and Biotechnology web site and will promote results there, for example in the News and Paper of the Month sections. We shall also promote it via chemical engineering contacts. MPW is also starting work on the second edition of his textbook How Proteins Work, which will incorporate any results of general interest.

Impact to industry will be pursued primarily through our industrial contacts, as being the best way of getting the message to the right people, and attracting support for our research. Here, we expect JWB to be an excellent ambassador, and to be energetic in promoting our research outcomes. The experience that he develops during the project in communicating to commercial organisations, and the contacts built up in this way, will be of great use to him in his future career. We shall maintain regular contact with the University's Research and Innovation Services team, who look after IP within the University. We realise the importance of building collaborations, which we see as important in developing the research further.

This research is relatively simple to explain to children and the general public, and it is fairly straightforward to explain why the research is worth doing. We shall therefore present our results in a number of ways at local Sheffield events, such as departmental open days, university open days, and the annual Discovery Night and Sheffield Science Week events. We have requested £500 to cover the costs of buying, testing and constructing robust demonstrations.