For further insight into the role of solvent in protein conformer stabilization, the structural and dynamic properties of protein ions in vacuo have been probed by hydrogen-deuterium exchange in a Fourier-transform mass spectrometer. Multiply charged ions generated by electrospray ionization of five proteins show exchange reactions with 2H20 at 10-7 torr (1 torr = 133.3 Pa) exhibiting pseudo-first-order kinetics. Gas-phase compactness of the S-S cross-linked RNase A relative to denatured S-derivatized RNase A is indicated by exchange of 35 and 135 hydrogen atoms, respectively. For pure cytochrome c ions, the existence of at least three distinct gaseous conformers is indicated by the substantially different values-52, 113, and 74 -ofreactive H atoms; the observation of these same values for ions of a number-2, 7, and 5, respectively-of different charge states indicates conformational insensitivity to coulombic forces. For each of these conformers, the compactness in vacuo indicated by these values corresponds directly to that of a known conformer structure in the solution from which the conformer ions are produced by electrospray. S-derivatized RNase A ions also exist as at least two gaseous conformers exchanging 50-140 H atoms. Gaseous conformer ions are isomerically stable for hours; removal of solvent greatly increases conformational rigidity. More specific ion-molecule reactions could provide further details of conformer structures.The relationship between the dynamic structure ofproteins in solution and their biological activity has been of longstanding research interest. Protein folding is probably the least well understood step in the sequence of transformations relating genetic information with its expression by protein function (1). Dramatic new ionization methods for mass spectrometry (MS) have made possible the formation of protein ions in the gas phase to measure molecular weight and primary sequence information (2-4), even on fmol samples (5, 6). Recent studies indicate that protein conformations in solution can affect the resulting charge distribution of the gaseous multiply charged ions formed by electrospray ionization (ESI) (7-9) and that even noncovalent complexes can survive ESI to form gaseous multiply charged ions (10)(11)(12)(13)(14)(15).Critical information concerning solvent effects on the conformation and dynamic properties of proteins has come from NMR (16) and from isotope-exchange experiments with 2H20 (17), including those before and during ESI/MS (18,19). With an activation energy of 17-20 kcal/mol (1 cal = 4.184 J) (17), the H/2H exchange rate depends on the pH (17), electrostatic effects (20), proximity of the solvent-accessible surface (21), and conformational flexibility with hydrogen bond cleavage and formation during local unfolding and folding (22). Studies ofgaseous proteins should help delineate the role of solvent in stabilizing protein conformations, but such previous studies have been mainly theoretical (23) because of the lack of experimental approaches. We repor...

Water is thought to play a dominant role in protein folding, yet gaseous multiply protonated proteins from which the water has been completely removed show hydrogen/deuterium (H/D) exchange behavior similar to that used to identify conformations in solution. Indicative of the gas-phase accessibility to D20, multiply-charged (6+ to 17+) cytochrome c cations exchange at six (or more) distinct levels of 64 to 173 out of 198 exchangeable H atoms, with the 132 H level found at charge values 8+ to 17+. Infrared laser heating and fast collisions can apparently induce ions to unfold to exchange at a higher distinct level, while chargestripping ions to lower charge values yields apparent folding as well as unfolding.Water is thought to be a key factor in the spontaneous folding of a protein into its bioactive conformation (1-5); "water interactions are of the essence in the function of real-life protein molecules" (6). However, a preliminary report (7) of the research described here gave hydrogen/deuterium (H/D) exchange evidence of at least three conformations of waterfree cytochrome c cations in the gas phase. These multiply protonated species were formed by electrospray ionization, introduced into the high-vacuum region of the Fouriertransform mass spectrometer, and allowed to exchange with D20 at 10-7 Torr (1 Torr = 133.3 Pa). In solution, the degree and sites of H/D exchange are well-established indicators of steric inaccessibility (8) and, thus, of protein conformation. To measure H/D exchange rates, these solution studies have used nuclear magnetic resonance (3, 9, 10), neutron diffraction (11), and mass spectrometry (MS), with MS employed to measure solution-phase H/D exchange rates (12-16), transient intermediates (17), and gaseous noncovalent complexes (18-21).For multiprotonated cytochrome c in the gas phase (7), H/D exchange exhibited pseudo-first-order kinetics for 7+ to 15+ ions. Three distinct exchange levels were dominant, with some charge values exhibiting two of three levels, but the level of highest exchange occurred at an intermediate charge value. Although in solution nonionized cytochrome c mimics this behavior, with the native structure first denatured and then reorganized into the molten globule state with increasing HI concentration (4, 5, 22), the importance of water for in vivo folding appears well established: "the likelihood of a protein refolding in the gas phase is exceedingly small" (23). We report here more definitive high-resolution MS data showing that multiply protonated equine cytochrome c ions undergo H/D exchange with D20 at six distinct levels, with conversion between these levels (presumably unfolding and folding) effected by infrared laser heating and high-velocity collisions or by reducing the number of charges on the ion. MATERIALS AND METHODSSolutions of 20 ,uM equine cytochrome c (Sigma) were electrosprayed [6+ to 9+ ions in pure aqueous solution and 10+ to 17+ ions in methanol/water/acetic acid, 76:22:2 (vol/vol)], and the resulting ions were transported by three rf-...

Department of Chemistry, Baker Laboratory, Cornell University, Ithaca, New York, USA Mass spectrometry instrumentation providing unit resolution and lo-ppm mass accuracy for molecules larger than 10 kDa was first reported in 1991. This instrumentation has now been improved with a 6.2-T magnet replacing that of 2.8 T, a more efficient vacuum system, ion injection with controlled ion kinetic energies, accumulated ion trapping with an open-cylindrical ion cell, acquisition of 2M data points, and updated electrospray apparatus. The resulting capabilities include resolving power of 5 × 10(5) for a 29-kDa protein, less than l-ppm mass measuring error, and dissociation of protein molecular ions to produce dozens of fragment ions whose exact masses can be identified from their mass-to-charge ratio values and isotopic peak spacing.

The total current and selected ion currents from the electrospray ionization (ES1) of 10(-5) M solutions of cocaine hydrochloride and deoxycytidine monophosphate (dCMP) monosodium salt in methanol and water solvents were compared in positive and negative ion modes, respectively, without and with SF6, gas as a discharge suppressant. The ESI onset voltages (Von), were the same for the positive and negative ion modes. The Von, for methanol was much lower than that for water and in agreement with the equation of D. P. H. Smith, who attributes the difference to the higher surface tension of water. The onset of electric discharge (Vdis) without SF6, occurred at lower capillary voltages for the negative relative to the positive ion mode for methanol; but Vdis is much higher than Von for methanol, and discharges do not interfere with ESI operation. For water, Von ≈ Vdis in the absence of SF6, in the negative ion mode, and ESI operation is impossible without SF6, discharge suppression. The discharge problem in the positive ion mode is less severe, but SF6, is still very useful. A dynamic range of 10 (-7)-10(-5) M was obtained by selected ion monitoring of [dCMP - H](-) at 4.5 and 20 μL/min. flows. Subpicomole detection limits for the nucleotide salt were obtained under these conditions.

Diacrylic acid (β‐acryloxypropionic acid) forms spontaneously in bulk acrylic acid by a Michael‐addition reaction. Fresh samples of acrylic acid containing 0.08–2.9% (m/m) levels of water were stored at constant temperatures of 15, 25, 35, and 45°C up to 65 days. An empirical equation to predict diacrylic acid formation was generated from the data.