Rebecca A. Jockusch

Sugars in the gas phase. Spectroscopy, conformation, hydration, co-operativity and selectivity

The functional importance of carbohydrates in biological processes, particularly those involving specific molecular recognition, is immense. Characterizing the three-dimensional structures of carbohydrates and glycoconjugates and their interactions with other molecules, particularly the ubiquitous solvent, water, are key starting points on the road towards the understanding of these processes. The review introduces a new strategy, combining electronic and vibrational spectroscopy of mass-selected carbohydrate molecules and their hydrated (and also protonated) complexes, conducted under molecular beam conditions, with ab initio computation. Its early successes have revealed a uniquely powerful means of characterizing carbohydrate conformations and hydrated structures, the hydrogen-bonded networks they support (or which support them) and the specificity of their interactions with other molecules. The new information, obtained in the gas phase, complements that provided by more ‘traditional’ condensed phase methods such as NMR, X-ray diffraction, molecular mechanics and molecular dynamics calculations. The review concludes with a vision of the challenges and opportunities offered by applications of molecular beam spectroscopy and their relevance in a biological context. Contents PAGE 1.  Preamble 490 2. Sweetness and light: Sugars in the gas phase 492 3. Experimental and computational strategies 495 4. The conformational landscapes of some key monosaccharides: glucose, galactose, mannose, fucose and xylose 498 4.1. Notation 498 4.2. Glucose, galactose and mannose 499 4.3. Fucose and xylose 503 5. Probing the glycosidic linkage: lactose and glycan ‘building blocks’ 504 5.1. Notation 506 5.2. Lactose 506 5.3. Mannose disaccharides 508 6. Adding water to sugar: hydrogen-bonding, co-operativity and selectivity 512 6.1. Notation 512 6.2. Mono-hydrated complexes: glucose, galactose and mannose 512 6.3. Co-operativity and conformational selectivity 516 6.4. Mono-hydrated complexes: xylose and fucose 519 6.5. Some concluding remarks 521 7. Using sugars: imino sugars and peptide mimics 522 7.1. Sugar mimics: imino sugars 522 7.2. Mimicking peptide secondary structure: carbopeptoids 524 8. Challenges and opportunities 527 Acknowledgements 529 References 530

Deactivation Pathways of an Isolated Green Fluorescent Protein Model Chromophore Studied by Electronic Action Spectroscopy

The mechanism of fluorescence and fluorescence quenching of the green fluorescent protein (GFP) is not well-understood. To gain insight into the effect of the surrounding protein on the chromophore buried at its center, the intrinsic electronic absorption and deactivation pathways of a gaseous model chromophore, p-hydroxybenzylidene-2,3-dimethylimidazolone (HBDI) were investigated. No fluorescence from photoactivated gaseous HBDI(-) was detected in the range 480-1100 nm, in line with the ultrafast rate of internal conversion of HBDI(-) in solution. Two different gas-phase deactivation pathways were found: photofragmentation and electron photodetachment. Electronic action spectra for each deactivation pathway were constructed by monitoring the disappearance of HBDI(-) and appearance of product ions as a function of excitation wavelength. The action spectra measured for each pathway are distinct, with electron photodetachment being strongly favored at higher photon energies. The combined (total) gas-phase action spectrum has a band origin at 482.5 nm (23340 cm(-1)) and covers a broad spectral range, 390-510 nm. This extended gas-phase action spectrum exhibits vibronic activity that matches well with the results of previous cold condensed-phase experiments and high-level in vacuo computations, with features evident at +550, +1500, and +2800 cm(-1) with respect to the band origin.

Fluorescence resonance energy transfer in gaseous, mass-selected polyproline peptides.

Despite the many successes of mass spectrometry in the analysis of biological samples, the need to better understand the correlation between condensed-phase properties and those of electrospray species remains. In particular, the link between structures in the condensed phase and in the gaseous environment of the mass spectrometer is still elusive. Here, we show that fluorescence resonance energy transfer (FRET) can be used to probe the conformations of gaseous biopolymers which are formed by electrospray ionization (ESI) and manipulated in a quadrupole ion trap mass spectrometer. A rhodamine dye pair suitable for gas-phase FRET is characterized. Both steady state spectra and lifetime measurements are used to monitor energy transfer in a series of dye-labeled polyproline-based peptides. FRET efficiency is explored as a function of peptide chain length and charge state. For the peptide with eight proline repeats, virtually complete energy transfer is observed. For the peptide with 14 proline repeats, energy transfer decreases as the charge state increases, consistent with Coulomb repulsion induced elongation of the peptide backbone. FRET measurements of the longest peptide examined, which has 20 proline repeats, indicates that the peptide adopts a bent configuration. Evidence for multiple conformations present within the ensemble of trapped ions is provided by fluorescence lifetime measurements. Gas-phase FRET measurements promise to be a new route to probe the conformations of large gaseous ions.

Gas-Phase Fluorescence Excitation and Emission Spectroscopy of Three Xanthene Dyes (Rhodamine 575, Rhodamine 590 and Rhodamine 6G) in a Quadrupole Ion Trap Mass Spectrometer

The gas-phase fluorescence excitation, emission and photodissociation characteristics of three xanthene dyes (rhodamine 575, rhodamine 590, and rhodamine 6G) have been investigated in a quadrupole ion trap mass spectrometer. Measured gas-phase excitation and dispersed emission spectra are compared with solution-phase spectra and computations. The excitation and emission maxima for all three protonated dyes lie at higher energy in the gas phase than in solution. The measured Stokes shifts are significantly smaller for the isolated gaseous ions than the solvated ions. Laser power-dependence measurements indicate that absorption of multiple photons is required for photodissociation. Redshifts and broadening of the dispersed fluorescence spectra at high excitation laser power provide evidence of gradual heating of the ion population, pointing to a mechanism of sequential multiple-photon activation through absorption/emission cycling. The relative brightness in the gas phase follows the order R575(1.00) < R590(1.15) < R6G(1.29). Fluorescence emission from several mass-selected product ions has been measured.

Sugars in the gas phase

The gas-phase conformation of the model glycoside, phenyl β-D-galactopyranoside (phe-β-D-gal) was examined using a combination of resonant two-photon ionization (R2PI) and resonant ion-dip infrared spectroscopy (RIDIRS) in tandem with electronic structure theory calculations. A single conformer, in which the hydroxy methyl is in a gauche+ orientation, is predominant in the free-jet expansion. This conformer is the analogue of the lowest-energy conformer found in a previous study of phenyl β-D-glucopyranoside. A minor second conformer has also been identified. Other weak bands in the R2PI spectrum have been attributed to hot bands though it is possible that other conformers are also present at low abundance. The dominant conformer identified in this work has the same (gauche+) orientation of the hydroxy methyl group as the major rotamer identified in solution by NMR spectroscopy.

Moving in on the Action: An Experimental Comparison of Fluorescence Excitation and Photodissociation Action Spectroscopy

Photodissociation action spectroscopy is often used as a proxy for measuring gas-phase absorption spectra of ions in a mass spectrometer. Although the potential discrepancy between linear optical and photodissociation spectra is generally acknowledged, direct experimental comparisons are lacking. In this work, we use a quadrupole ion trap that has been modified to enable both photodissociation and laser-induced fluorescence to assess how closely the visible photodissociation action spectrum of a fluorescent dye reflects its fluorescence excitation spectrum. Our results show the photodissociation action spectrum of gaseous rhodamine 110 is both substantially narrower and slightly red-shifted (∼120 cm(-1)) compared to its fluorescence excitation spectrum. Power dependence measurements reveal that the photodissociation of rhodamine 110 requires, on average, the absorption of three photons whereas fluorescence is a single-photon process. These differing power dependences are the key to interpreting the differences in the measured spectra. The experimental results provide much-needed quantification and insight into the differences between action spectra and linear optical spectra, and emphasize the utility of fluorescence excitation spectra to provide a more reliable benchmark for comparison with theory.

Gas-Phase FRET Efficiency Measurements To Probe the Conformation of Mass-Selected Proteins

Electrospray ionization and mass spectrometry have revolutionized the chemical analysis of biological molecules, including proteins. However, the correspondence between a proteins native structure and its structure in the mass spectrometer (where it is gaseous) remains unclear. Here, we show that fluorescence (Förster) resonance energy transfer (FRET) measurements combined with mass spectrometry provides intramolecular distance constraints in gaseous, ionized proteins. Using an experimental setup which combines trapping mass spectrometry and laser-induced fluorescence spectroscopy, the structure of a fluorescently labeled mutant variant of the protein GB1 was probed as a function of charge state. Steady-state fluorescence emission spectra and time-resolved donor fluorescence measurements of mass-selected GB1 show a marked decrease in the FRET efficiency with increasing number of charges on the gaseous protein, which suggests a Coulombically driven unfolding and expansion of its structure. This lies in stark contrast to the pH stability of GB1 in solution. Comparison with solution-phase single-molecule FRET measurements show lower FRET efficiency for all charge states of the gaseous protein examined, indicating that the ensemble of conformations present in the gas phase is, on average, more expanded than the native form. These results represent the first FRET measurements on a mass-selected protein and illustrate the utility of FRET for obtaining a new kind of structural information for large, desolvated biomolecules.

Understanding Photophysical Effects of Cucurbituril Encapsulation: A Model Study with Acridine Orange in the Gas Phase

Encapsulation of dyes by cucurbituril macrocycles has proven profitable as a strategy to alter fluorescence characteristics in useful ways. Encapsulation generally results in longer fluorescence lifetimes, enhanced brightness, and solvatochromic effects not normally seen in the condensed phase. These effects have been attributed variously to both the removal of interactions with solvent molecules and to the confined environment of extremely low polarizability provided by the cucurbituril interior. It is difficult to disentangle these effects in solution. Here, we present results from gas-phase experiments designed to separate these effects, using cucurbit[7]uril (CB7), and the cationic dye acridine orange (AOH(+)) as a probe. Fluorescence properties of gaseous AOH(+) are compared with those of the gaseous AOH(+)-CB7 complex and with the properties of the dye and complex in aqueous solution. The dependence on the local environment of several spectroscopic properties is discussed, including the fluorescence excitation and emission maxima, the size of the Stokes shift, fluorescence lifetime and relative brightness. An understanding of the modulation of fluorescence properties by the local environment, such as that promoted by this work, will aid in the rational design of improved fluorophores and fluorescent sensors.

Efavirenz Binding Site in HIV-1 Reverse Transcriptase Monomers

Efavirenz (EFV) is a potent nonnucleoside reverse transcriptase inhibitor (NNRTI) used in the treatment of AIDS. NNRTIs bind in a hydrophobic pocket located in the p66 subunit of reverse transcriptase (RT), which is not present in crystal structures of RT without an inhibitor. Recent studies showed that monomeric forms of the p66 and p51 subunits bind efavirenz with micromolar affinity. The effect of efavirenz on the solution conformations of p66 and p51 monomers was studied by hydrogen-deuterium exchange mass spectrometry (HXMS) and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). HXMS data reveal that five peptides, four of which contain efavirenz contact residues seen in the crystal structure of the RT-EFV complex, exhibit a reduced level of exchange in monomer-EFV complexes. Moreover, peptide 232-246 undergoes slow cooperative unfolding-refolding in the bound monomers, but at a rate much slower than that observed in the p66 subunit of the RT heterodimer [Seckler, J. M., Howard, K. J., Barkley, M. D., and Wintrode, P. L. (2009) Biochemistry 48, 7646-7655]. These results suggest that the efavirenz binding site on p66 and p51 monomers is similar to the NNRTI binding pocket in the p66 subunit of RT. Nanoelectrospray ionization FT-ICR mass spectra indicate that the intact monomers each have (at least) two different conformations. In the presence of efavirenz, the mass spectra change significantly and suggest that p51 adopts a single, more compact conformation, whereas p66 undergoes facile, electrospray-induced cleavage. The population shift is consistent with a selected-fit binding mechanism.

Ligand Migration in the Gaseous Insulin-CB7 Complex—A Cautionary Tale About the Use of ECD-MS for Ligand Binding Site Determination

Knowledge of the structure of protein–ligand complexes can aid in understanding their roles within complex biological processes. Here we use electrospray ionization (ESI) coupled to a Fourier transform ion cyclotron resonance mass spectrometer to investigate the noncovalent binding of the macrocycle cucurbit[7]uril (CB7) to bovine insulin. Recent condensed-phase experiments (Chinai et al., J. Am. Chem. Soc. 133:8810–8813, 2011) indicate that CB7 binds selectively to the N-terminal phenylalanine of the insulin B-chain. Competition experiments employing ESI mass spectrometry to assess complex formation between CB7 and wild type insulin B-chain vs. a mutant B-chain, confirm that the N-terminal phenylalanine plays in important role in solution-phase binding. However, analysis of fragment ions produced by electron capture dissociation (ECD) of CB7 complexed to intact insulin and to the insulin B-chain suggests a different picture. The apparent gas-phase binding site, as identified by the ECD, lies further along the insulin B-chain. Together, these studies thus indicate that the CB7 ligand migrates in the ESI mass spectrometry analysis. Migration is likely aided by the presence of additional interactions between CB7 and the insulin B-chain, which are not observed in the crystal structure. While this conformational difference may result simply from the removal of solvent and addition of excess protons by the ESI, we propose that the migration may be enhanced by charge reduction during the ECD process itself because ion-dipole interactions are key to CB7 binding. The results of this study caution against using ECD-MS as a stand-alone structural probe for the determination of solution-phase binding sites.