Zwitterions - Bradykinin

    The singly protonated nonapeptide bradykinin (BK) with the sequence arg-pro-pro-gly-phe-ser-pro-phe-arg is ideally set up to form a salt bridge in the gas phase with both arginine side chains protonated and the C-terminus deprotonated. However, from cross section experiments in comparison with calculations the zwitterion hypothesis can neither be confirmed nor rejected as both salt bridge and charge solvation structures cover the same range of cross sections.[17]
     The large set of model structures obtained as part of the cross section project was also used to attempt to answer the question: Does gas-phase H/D-exchange data provide information about the gas-phase structure of ions with labile hydrogen atoms? To address this question we developed a simple model to predict the H/D-exchange kinetics for a set of given model structures. Using our bradykinin molecular mechanics structures obtained both for zwitterions and charge-solvation structures, we found that a set of low-energy zwitterion structures matched the experimental data obtained by others [18] far better than a corresponding set of non-zwitterion structures.[14]
    The main features of the model we developed include:[14]

• D₂O samples entire peptide surface
• peptide conformation does not change upon addition of D₂O
• H/D-exchange occurs by a "relay" mechanism,[19] hence
  • the protonated site –AH⁺ has to be accessible to D₂O
  • a "basic site" -B: has to be accessible to D₂O
  • -B: must be close to –AH⁺.
• peptide conformation does not change during collision time
• peptide conformation may change on experimental time scale
  

    To check on some of the assumptions above we ran molecular dynamics calculations on the (BK + H + D₂O)⁺ system. We found that D₂O does move along the peptide surface, but that it does prefer to hang out at certain locations on the peptide surface for extended periods of time. Two of those locations, both near a protonated arginine side chain, are indicated in the figure below for the example of a bradykinin salt bridge structure.

    Recent hydration experiments carried out on protonated bradykinin yield a pattern of water binding energies that is very different from the protonated peptide LHRH, a decapeptide which cannot form a salt bridge structure because it does not have any acidic site necessary for deprotonation in the case of a zwitterion. Experimental water binding energies for the first four water molecules adding to protonated bradykinin are nearly identical, whereas for protonated LHRH the first water molecule is more strongly bound than the second, which is in turn more strongly bound than the third and fourth water molecule (see table above). Molecular mechanics studies on these two systems including one to four water molecules indicate quite different hydration properties for the two peptides in line with the different patterns of hydration energies. For protonated LHRH the water molecules solvate the peptide surface rather evenly with a slight preference for adding the first water molecule to the protonated site. For protonated bradykinin, on the other hand, water molecules prefer to form a loop bound on both sides to the salt bridge: -COO^-···H₂O···H₂O···H₂O···H⁺Arg-.
It is reasonable to assume that inserting water molecules into the loop lowers the system energy by a constant amount per water molecule inserted, while the bradykinin conformation does not have to undergo any major changes. This would explain perfectly well, why the first few water molecules have nearly identical binding energies. Note, that flipping the bonds in the loop converts the zwitterion into a charge solvation structure:
-COOH···OH₂···OH₂···OH₂···Arg-

Notes

Experiment: The original cross section experiments were carried out on our MALDI ion mobility instrument. The hydration experiments were carried out on our ESI ion mobility instrument. Also see the Peptide Hydration research project.

Model structures: Model structures were obtained by a molecular-mechanics-based simulated annealing protocol.