One of the most important- and least understood- effects dealing with biomolecular structure concerns the effects of salts on protein conformation. The famous Hofmeister Series for ions that either 'salt-in' or 'salt-out' proteins is well known, but the mechanism through which the ions act is controversial and probably involves not one mechanism but different ones under different circumstances.
In an interesting single-author JACS paper, Joachim Dzubiella studied the effects of different salts of sodium and potassium on the structure of alpha helices in solution. Even something as common and widely studied as the alpha helix is still an enigma. For example, the simple question "What contributes to the stability of an alpha helix?" is controversial and not fully answered yet. In this context I will refer the reader to an excellent perspective written by Robert Baldwin at Stanford that tries to answer the rather simple question: "How much energetic stability does a peptide bond in a helix contribute?". Baldwin looks at two approaches to understand the problem. One is the 'hydrogen bond inventory' approach which simply lists the bonds broken and formed on each side when an amide group desolvates and forms a peptide bond. Based on this approach, the mean figure for peptide h-bond energy has been estimated as 0.9 kcal/mol/h-bond. Even though this quantity is small, a 100 residue protein where 70 residues form hydrogen bonds is clearly going to lead to a very substantial net stabilization. The second approach that Baldwin considers is the electrostatic solvation enthalpy or free-energy method, where one uses the Born equation to estimate the strength of a h-bond. Using this approach Baldwin gets a very different answer- 2.5 kcal/mol. Clearly there is still some way to go toward estimating how much peptide h-bonds contribute to stability. One important factor not considered by Baldwin is the entropy of the water. Another important factor that he does consider is the preferential desolvation for helix formation that depends on the exact residues involved. We have ourselves encountered desolvation issues in continuing work on amyloid beta-sheets.
But back to Dzubiella's paper. Dzubiella uses MD simulations to study the dynamics of helix-salt interaction. He considers helices where a i---> (i+4) salt bridge between the side chains of a glutamate and lysine has stabilized the conformation. He looks at which salts stabilize helices and which ones destabilize them. From these detailed simulations he gains some valuable insight into the radically different behavior of rather similar ions. For example, K+ ions are much less able to destabilize helices than Na+ ions. This is due to preferential interaction of carboxylate groups involved in salt-bridge formation by Na+. Due to its smaller size, Na+ is better able to interact with carboxylates than K+.
However, we have to remember that Na+ or K+ or any of the other ions have to compete with water when interacting with amino acids in the peptide. Water is in great excess and water also efficiently interacts with carboxylates (1). The MD simulations reveal that a curious and unexpected helper comes to the aid of the Na+ ions- I- ions. Iodide interestingly interacts with the non-polar parts of the peptide, thus "clearing" water away and paving the way for Na+ to access the carboxylates and carbonyls. This unexpected observation again sheds light on the different properties of iodine compared to the rest of the halogens (2). Iodide is much bigger, has a diffuse charge and is therefore much more polarizable. Apparently it is so electronically watered down that even carbon thinks it is harmless and can preferentially interact with it.
This curious observation tells us that we know less about the elements than we think. From the observation of weak hydrogen bonds and halogen bonds to the unexpected non-polarity of iodide, surprises await us in the realm of biomolecular structure and indeed in all of chemistry. It is also thanks to tools like MD that we can now gain insights into the details of such molecular interaction.
Notes: (1) In fact water can interact so well that it might steal a few h-bonds from the peptide and destabilize the helix. That's why trifluoroethanol (TFE) or hexafluoroacetone are so good at stabilizing helices (these lead to "Teflon-coated peptides"), because the fluorine cannot steal h-bonds from the peptide backbone.
(2) For example iodide most efficiently forms halogen bonds with oxygen, a phenomenon now well-accepted.
References: Joachim Dzubiella (2008). Salt-Specific Stability and Denaturation of a Short Salt-Bridge-Forming α-Helix Journal of the American Chemical Society DOI: 10.1021/ja805562g
R. L. Baldwin (2003). In Search of the Energetic Role of Peptide Hydrogen Bonds Journal of Biological Chemistry, 278 (20), 17581-17588 DOI: 10.1074/jbc.X200009200