Christian Bleiholder

Ion mobility–mass spectrometry reveals a conformational conversion from random assembly to β-sheet in amyloid fibril formation

Amyloid cascades that lead to peptide β-sheet fibrils and plaques are central to many important diseases. Recently, intermediate assemblies of these cascades were identified as the toxic agents that interact with cellular machinery. The location and cause of the transformation from a natively unstructured assembly to the β-sheet oligomers found in all fibrils is important in understanding disease onset and the development of therapeutic agents. Largely, research on this early oligomeric region was unsuccessful because all the traditional techniques measure only the average oligomer properties of the ensemble. We utilized ion-mobility methods to deduce the peptide self-assembly mechanism and examined a series of amyloid-forming peptides clipped from larger peptides or proteins associated with disease. We provide unambiguous evidence for structural transitions in each of these fibril-forming peptide systems and establish the potential of this method for the development of therapeutic agents and drug evaluation.

Sequence-scrambling fragmentation pathways of protonated peptides.

The gas-phase structures and fragmentation pathways of the N-terminal b and a fragments of YAGFL-NH(2), AGLFY-NH(2), GFLYA-NH(2), FLYAG-NH(2), and LYAGF-NH(2) were investigated using collision-induced dissociation (CID) and detailed molecular mechanics and density functional theory (DFT) calculations. Our combined experimental and theoretical approach allows probing of the scrambling and rearrangement reactions that take place in CID of b and a ions. It is shown that low-energy CID of the b(5) fragments of the above peptides produces nearly the same dissociation patterns. Furthermore, CID of protonated cyclo-(YAGFL) generates the same fragments with nearly identical ion abundances when similar experimental conditions are applied. This suggests that rapid cyclization of the primarily linear b(5) ions takes place and that the CID spectrum is indeed determined by the fragmentation behavior of the cyclic isomer. This can open up at various amide bonds, and its fragmentation behavior can be understood only by assuming a multitude of fragmenting linear structures. Our computational results fully support this cyclization-reopening mechanism by showing that protonated cyclo-(YAGFL) is energetically favored over the linear b(5) isomers. Furthermore, the cyclization-reopening transition structures are energetically less demanding than those of conventional bond-breaking reactions, allowing fast interconversion among the cyclic and linear isomers. This chemistry can lead in principle to complete loss of sequence information upon CID, as documented for the b(5) ion of FLYAG-NH(2). CID of the a(5) ions of the above peptides produces fragment ion distributions that can be explained by assuming b-type scrambling of their parent population and a --> a*-type rearrangement pathways ( Vachet , R. W. , Bishop , B. M. , Erickson , B. W. , and Glish , G. L. J. Am. Chem. Soc. 1997, 119, 5481 ). While a ions easily undergo cyclization, the resulting macrocycle predominantly reopens to regenerate the original linear structure. Computational data indicate that the a --> a*-type rearrangement pathways of the linear a isomers involve post-cleavage proton-bound dimer intermediates in which the fragments reassociate and the originally C-terminal fragment is transferred to the N-terminus.

Protomers of Benzocaine: Solvent and Permittivity Dependence

The immediate environment of a molecule can have a profound influence on its properties. Benzocaine, the ethyl ester of para-aminobenzoic acid that finds an application as a local anesthetic, is found to adopt in its protonated form at least two populations of distinct structures in the gas phase, and their relative intensities strongly depend on the properties of the solvent used in the electrospray ionization process. Here, we combine IR-vibrational spectroscopy with ion mobility-mass spectrometry to yield gas-phase IR spectra of simultaneously m/z and drift-time-resolved species of benzocaine. The results allow for an unambiguous identification of two protomeric species: the N- and O-protonated forms. Density functional theory calculations link these structures to the most stable solution and gas-phase structures, respectively, with the electric properties of the surrounding medium being the main determinant for the preferred protonation site. The fact that the N-protonated form of benzocaine can be found in the gas phase is owed to kinetic trapping of the solution-phase structure during transfer into the experimental setup. These observations confirm earlier studies on similar molecules where N- and O-protonation have been suggested.

Ion mobility spectrometry reveals the mechanism of amyloid formation of Aβ(25-35) and its modulation by inhibitors at the molecular level: epigallocatechin gallate and scyllo-inositol.

Amyloid cascades leading to peptide β-sheet fibrils and plaques are central to many important diseases. Recently, intermediate assemblies of these cascades were identified as the toxic agents that interact with the cellular machinery. The relationship between the transformation from natively unstructured assembly to the β-sheet oligomers to disease is important in understanding disease onset and the development of therapeutic agents. Research on this early oligomeric region has largely been unsuccessful since traditional techniques measure only ensemble average oligomer properties. Here, ion mobility methods are utilized to deduce the modulation of peptide self-assembly pathways in the amyloid-β protein fragment Aβ(25-35) by two amyloid inhibitors (epigallocatechin gallate and scyllo-inositol) that are currently in clinical trials for Alzheimers Disease. We provide evidence that suppression of β-extended oligomers from the onset of the conversion into β-oligomer conformations is essential for effective attenuation of β-structured amyloid oligomeric species often associated with oligomer toxicity. Furthermore, we demonstrate the ease with which ion mobility spectrometry-mass spectrometry can guide the development of therapeutic agents and drug evaluation by providing molecular level insight into the amyloid formation process and its modulation by small molecule assembly modulators.

Factors Contributing to the Collision Cross Section of Polyatomic Ions in the Kilodalton to Gigadalton Range: Application to Ion Mobility Measurements

The projected superposition approximation (PSA) method was used to theoretically evaluate the factors contributing to the cross section measured in ion mobility experiments and to study how the significance of these factors varies with ion size from diglycine to a 1 μm oil droplet. Thousands of PSA calculations for ∼400 different molecules in the temperature range from 80 to 700 K revealed that the molecular framework made up of atomic hard spheres is, as expected, a major component of the cross section. However, the ion-buffer gas interaction is almost equally important for very small peptides, and although its significance decreases with increasing ion size, interaction is still a factor for megadalton ions. An additional major factor is the ion shape: Fully convex ions drifting in a buffer gas have a minimal frictional resisting force, whereas the resisting force increases with degree of ion surface concaveness. This added resistance is small for peptides and larger for proteins and increases the ion mobility cross section from 0 to greater than 40%. The proteins with the highest degree of concaveness reach a shape-effected friction similar to, and sometimes larger than that of, macroscopic particles such as oil droplets. In summary, our results suggest that the transition from nanoparticle (with Lennard-Jones-like interaction with the buffer gas) to macroscopic particle (with hard sphere-like interaction) occurs at ∼1 GDa.

Molecular Structures and Ion Mobility Cross Sections: Analysis of the Effects of He and N2 Buffer Gas

An empirically observed correlation between ion mobility cross sections in helium and nitrogen buffer gases was examined as a function of temperature, molecular size, and shape. Experimental cross sections were determined for tetraglycine, bradykinin, angiotensin 2, melittin, and ubiquitin at 300 K and in the range from 80 to 550 K on home-built instruments and calculated by the projection superposition approximation (PSA) method. The PSA was also used to predict cross sections for larger systems such as human pancreatic alpha-amylase, concanavalin, Pichia pastoris lysyl oxidase, and Klebsiella pneumoniae acetolactate synthase. The data show that the ratio of cross sections in helium and nitrogen depends significantly on the temperature of the buffer gas as well as the size and shape of the analyte ion. Therefore, the analysis of the data indicates that a simple formula that seeks to quantitatively relate the momentum transfer cross sections observed in two distinct buffer gases lacks a sound physical basis.

Differentiation of CC vs CXC Chemokine Dimers with GAG Octasaccharide Binding Partners: An Ion Mobility Mass Spectrometry Approach

Chemokines, 8 kDa proteins implicated in leukocyte migration via oligomerization, bind to glycosaminoglycans (GAGs) during the inflammation response as a means to regulate chemokine migration. Structural characterization of chemokines non-covalently bound to GAGs provides physiologically meaningful data in regard to routine inmmunosurveillance and disease response. In order to analyze the structures resulting from the GAG:chemokine interaction, we employed ion mobility mass spectrometry (IMMS) to analyze monocyte chemoattractant protein-1 (MCP-1), a CC chemokine, and interleukin-8 (IL-8), a CXC chemokine, along with their individual interactions with GAG heparin octasaccharides. We show that MCP-1 and IL-8 are physiologically present as a dimer, with MCP-1 having two variants of its dimeric form and IL-8 having only one. We also show that the MCP-1 dimer adopts two conformations, one extended and one compact, when bound to a dodecasulfated heparin octasaccharide. Binding of MCP-1 to heparin octasaccharide isomers of varying sulfation patterns results in similar arrival time distribution values, which suggests minimal distinguishing features among the resultant complexes. Additionally, tandem mass spectrometry (MS/MS) showed that the binding of MCP-1 to a heparin octasaccharide has different dissociation patterns when compared with the corresponding IL-8 bound dimer. Overall, IMMS and MS/MS were used to better define the structural tendencies and differences associated with CC and CXC dimers when associated with GAG octasaccharides.

Factors That Drive Peptide Assembly and Fibril Formation: Experimental and Theoretical Analysis of Sup35 NNQQNY Mutants

Residue mutations have substantial effects on aggregation kinetics and propensities of amyloid peptides and their aggregate morphologies. Such effects are attributed to conformational transitions accessed by various types of oligomers such as steric zipper or single β-sheet. We have studied the aggregation propensities of six NNQQNY mutants: NVVVVY, NNVVNV, NNVVNY, VIQVVY, NVVQIY, and NVQVVY in water using a combination of ion-mobility mass spectrometry, transmission electron microscopy, atomic force microscopy, and all-atom molecular dynamics simulations. Our data show a strong correlation between the tendency to form early β-sheet oligomers and the subsequent aggregation propensity. Our molecular dynamics simulations indicate that the stability of a steric zipper structure can enhance the propensity for fibril formation. Such stability can be attained by either hydrophobic interactions in the mutant peptide or polar side-chain interdigitations in the wild-type peptide. The overall results display only modest agreement with the aggregation propensity prediction methods such as PASTA, Zyggregator, and RosettaProfile, suggesting the need for better parametrization and model peptides for these algorithms.

From Noncovalent Chalcogen–Chalcogen Interactions to Supramolecular Aggregates: Experiments and Calculations

This review considers noncovalent bonds between divalent chalcogen centers. In the first part we present X-ray data taken from the solid state structures of dimethyl- and diphenyl-dichalcogenides as well as oligoalkynes kept by alkyl-sulfur, -selenium, and -tellurium groups. Furthermore, we analyzed the solid state structures of medium sized (12-24 ring size) selenium coronands and medium to large rings with alkyne and alkene units between two chalcogen centers. The crystal structures of the cyclic structures revealed columnar stacks with close contacts between neighboring rings via noncovalent interactions between the chalcogen centers. To get larger space within the cavities, rings with diyne units between the chalcogen centers were used. These molecules showed channel-like structures in the solid state. The flexibility of the rings permits inclusion of guest molecules such as five-membered heterocycles and aromatic six-membered rings. In the second part we discuss the results of quantum chemical calculations. To treat properly the noncovalent bonding between chalcogens, we use diffuse augmented split valence basis sets in combination with electron correlation methods. Our model substances were 16 dimers consisting of two Me-X-Me (X = O, S, Se, Te) pairs and dimers of Me-X-Me/Me-X-CN (X = O, S, Se, Te) pairs. The calculations show the anticipated increase of the interaction energy from (Me-O-Me)2 (-2.15 kcal/mol) to (Me-O-Me/Me-Te-CN) (-6.59 kcal/mol). An analysis by the NBO method reveals that in the case of the chalcogen centers O and S the hydrogen bridges between the molecules dominate. However, in the case of Se and Te the major bonding between the pairs originates from dispersion forces between the chalcogen centers. It varies between -1.7 and -4.0 kcal/mol.

The Solution Assembly of Biological Molecules Using Ion Mobility Methods: From Amino Acids to Amyloid β-Protein

Ion mobility spectrometry-mass spectrometry (IMS-MS) methods are increasingly used to study noncovalent assemblies of peptides and proteins. This review focuses on the noncovalent self-assembly of amino acids and peptides, systems at the heart of the amyloid process that play a central role in a number of devastating diseases. Three different systems are discussed in detail: the 42-residue peptide amyloid-β42 implicated in the etiology of Alzheimers disease, several amyloid-forming peptides with 6-11 residues, and the assembly of individual amino acids. We also discuss from a more fundamental perspective the processes that determine how quickly proteins and their assemblies denature when the analyte ion has been stripped of its solvent in an IMS-MS measurement and how to soften the measurement so that biologically meaningful data can be recorded.