A top-down approach to protein structure studies using chemical cross-linking and Fourier transform mass spectrometry

In a preliminary communication we described a top-down approach to the determination of chemical cross-link location in proteins using Fourier transform mass spectrometry (FT-MS). We have since extended the approach to use a series of homobifunctional cross-linkers with the same reactive functional groups, but different cross-linker arm lengths. Correlating cross-linking data across a series of related linkers allows the distance constraint derived from a cross-link between two reactive side chains to be determined more accurately and increases the confidence in the assignment of the cross-links. In ubiquitin, there are seven lysines with primary amino groups and the amino terminus. Disuccinimidyl suberate (DSS, cross-linker arm length = 11.4 A), disuccinimidyl glutarate (DSG, cross-linker arm length = 7.5 A) and disuccinimidyl tartrate (DST, cross- linker arm length = 5.8 A) are homobifunctional cross-linking reagents that react specifically with primary amines. Using tandem mass spectrometry (MS/MS) on the singly, internally cross-linked precursor ion of ubiquitin, we found cross-links with DSS and DSG between the amino terminus and Lys 6, between Lys 6 and Lys 11, and between Lys 63 and Lys 48. Using disuccinimidyl tartrate (DST), the shortest cross-linker in the series, only the cross-links between the amino terminus and Lys 6, and between Lys 6 and Lys 11 were observed. The observed cross-links are consistent with the crystal structure of ubiquitin, if the lysine side chains and the amino terminus are assumed to have considerable flexibility. In a separate study, we probed the reactivity of the primary amino groups in ubiquitin using the amino acetylating reagent, N-hydroxy succinimidyl acetate (NHSAc), and a top-down approach to localize the acetylated lysine residues. The reactivity order obtained in that study (M1 approximate, equals K6 approximate, equals K48 approximate, equals K63) > K33 > K11 > (K27, K29), shows that the cross-link first formed in ubiquitin by reaction with DSS and DSG occurs between the most reactive residues.     

Conformational dependence of the intrinsic acidity of the aspartic acid residue sidechain in N-acetyl- -aspartic acid-N′-methylamide

The sidechain conformational potential energy hypersurfaces (PEHS) for the γ_L, β_L, α_L, and α_D backbone conformations of N-acetyl- -aspartate-N′-methylamide were generated. Of the 81 possible conformers initially expected for the aspartate residue, only seven were found after geometric optimizations at the B3LYP/6-31G(d) level of theory. No stable conformers could be located in the δ_L, _L, γ_D, δ_D, and _D backbone conformations. The ‘adiabatic’ deprotonation energies for the endo and exo forms of N-acetyl- -aspartic acid-N′-methylamide were calculated by comparing their optimized relative energies against those found for the seven stable conformers of N-acetyl- -aspartate-N′-methylamide. Sidechain conformational PEHSs were also generated for the estimation of ‘vertical’ deprotonation energies for both endo and exo forms of N-acetyl- -aspartic acid-N′-methylamide. All backbone–sidechain (N–H⁻O–C) and backbone–backbone (N–HO=C) hydrogen bond interactions were analyzed. A total of two backbone–backbone and four backbone–sidechain interactions were found for N-acetyl- -aspartate-N′-methylamide. The deprotonated sidechain of N-acetyl- -aspartate-N′-methylamide may allow the aspartyl residue to form strong hydrogen bond interactions (since it is negatively charged) which may be significant in such processes as protein–ligand recognition and ligand binding. As a primary example, the molecular geometry of the aspartyl residue may be important in peptide folding, such as that in the RGD tripeptide.     

Stereochemistry of [4 + 2] Cycloadditions of Perfluorinated Selenocarbonyls to 1, 3‐Dienes

In completely stereospecific [4+2] cycloadditions, the perfluorinated selenocarbonyls 1 and 2 react both with trans‐trans‐2, 4‐hexadiene and cis‐trans‐2, 4‐hexadiene to yield 3, 6‐dihydro‐cis‐3, 6‐dimethyl‐2H‐selenapyrans 3 , 4a and 4b . The observed stereoselectivity leads to the conclusion, that the [4+2] cycloaddition of perfluorinated selenocarbonyls follows a concerted pathway. An identical mixture of isomers was isolated when using the precursor for 2 , trimethylstannyl (pentafluoroethyl)selane, which reacts with both 1, 3‐dienes over several weeks to form a mixture of syn‐2‐fluoro‐3, 6‐dihydro‐cis‐3, 6‐dimethyl‐2‐trifluoromethyl‐2H‐selenapyran ( 4a ) and anti‐2‐fluoro‐3, 6‐dihydro‐cis‐3, 6‐dimethyl‐2‐trifluoromethyl‐2H‐selenapyran ( 4b ) in the same ratio as found for 2 , thus proving the intermediate formation of Se=C(F)CF₃ ( 2 ). Complex 2D NMR experiments were used to distinguish the isomers 4a and 4b and to assign the ¹H, ¹³C and ¹⁹F NMR data of the selenaheterocycles.     

Microphase separation of cationic poly(N-isopropylacrylamide) copolymers in water: Effect of the migration of charges

The structures of aqueous copolymer solutions have been examined through small angle neutron scattering. The copolymers contained mostly N-isopropylacrylamide (NIPAM) monomers. Poly (NIPAM) solutions have a lower critical solution temperature (LCST), above which the macromolecules separate from water. A small fraction of ionizable N,N-[(dimethylamino) propyl] methacrylamide (MADAP) monomers was introduced into the macromolecules. This had dramatic consequences on the solution behavior at temperatures above the LCST of PNIPAM, where phase separation would have been expected for the homopolymer. When all MADAP monomers were ionized, it was found that the solutions resisted the phase separation. At short spatial scales, the chains were collapsed but at large scales they formed branched aggregates that did not separate out of water. When only half of the MADAP monomers are ionized, the electrical charges were able to redistribute themselves along the chains. In this case, the rise in temperature caused a microphase separation where the electrical charges were relocated on a fraction of the chains that remained in solution.The other chains (or section of chains) formed large nodules of a polymer rich phase.     

Six-coordinate and five-coordinate Fe(II)(CN)(2)(CO)(x) thiolate complexes (x = 1, 2): synthetic advances for iron sites of [NiFe] hydrogenases

The dicyanodicarbonyliron(II) thiolate complexes trans,cis-[(CN)(2)(CO)(2)Fe(S,S-C-R)](-) (R = OEt (2), N(Et)(2) (3)) were prepared by the reaction of [Na][S-C(S)-R] and [Fe(CN)(2)(CO)(3)(Br)](-) (1). Complex 1 was obtained from oxidative addition of cyanogen bromide to [Fe(CN)(CO)(4)](-). In a similar fashion, reaction of complex 1 with [Na][S,O-C(5)H(4)N], and [Na][S,N-C(5)H(4)] produced the six-coordinate trans,cis-[(CN)(2)(CO)(2)Fe(S,O-C(5)H(4)N)](-) (6) and trans,cis-[(CN)(2)(CO)(2)Fe(S,N-C(5)H(4))](-) (7) individually. Photolysis of tetrahydrofuran (THF) solution of complexes 2, 3, and 7 under CO led to formation of the coordinatively unsaturated iron(II) dicyanocarbonyl thiolate compounds [(CN)(2)(CO)Fe(S,S-C-R)](-) (R = OEt (4), N(Et)(2) (5)) and [(CN)(2)(CO)Fe(S,N-C(5)H(4))](-) (8), respectively. The IR v(CN) stretching frequencies and patterns of complexes 4, 5, and 8 have unambiguously identified two CN(-) ligands occupying cis positions. In addition, density functional theory calculations suggest that the architecture of five-coordinate complexes 4, 5, and 8 with a vacant site trans to the CO ligand and two CN(-) ligands occupying cis positions serves as a conformational preference. Complexes 2, 3, and 7 were reobtained when the THF solution of complexes 4, 5, and 8 were exposed to CO atmosphere at 25 degrees C individually. Obviously, CO ligand can be reversibly bound to the Fe(II) site in these model compounds. Isotopic shift experiments demonstrated the lability of carbonyl ligands of complexes 2, 3, 4, 5, 7, and 8. Complexes [(CN)(2)(CO)Fe(S,S-C-R)](-) and NiA/NiC states [NiFe] hydrogenases from D. gigas exhibit a similar one-band pattern in the v(CO) region and two-band pattern in the v(CN) region individually, but in different positions, which may be accounted for by the distinct electronic effects between [S,S-C-R](-) and cysteine ligands. Also, the facile formations of five-coordinate complexes 4, 5, and 8 imply that the strong sigma-donor, weak pi-acceptor CN(-) ligands play a key role in creating/stabilizing five-coordinate iron(II) [(CN)(2)(CO)Fe(S,S-C-R)](-) complexes with a vacant coordination site trans to the CO ligand.     

Two-particle friction in a mesoscopic solvent

The effects of hydrodynamic interactions on the friction tensors for two particles in solution are studied. The particles have linear dimensions on nanometer scales and are either simple spherical particles interacting with the solvent through repulsive Lennard-Jones forces or are composite cluster particles whose atomic components interact with the solvent through repulsive Lennard-Jones forces. The solvent dynamics is modeled at a mesoscopic level through multiparticle collisions that conserve mass, momentum, and energy. The dependence of the two-particle relative friction tensors on the interparticle separation indicates the importance of hydrodynamic interactions for these nanoparticles.     

Towards two-dimensional electrophoresis mapping of the cerebrospinal fluid proteome from a single individual

We test the ability of state-of-the-art two-dimensional electrophoresis (2-DE) technology to enable the proteome mapping of ante mortem cerebrospinal fluid (CSF) from a single individual. Using the sensitive technologies of a fluorescent protein stain and fluorescence laser scanning of 2-DE gels, combined with matrix-assisted laser desorption/ionization-time of flight/time of flight-mass spectrometry (MALDI-TOF/TOF-MS) for protein identification, a highly detailed 2-DE map of the CSF proteome was created. The 2-DE map contains 600 identified spots representing 82 different proteins. Of the 82 proteins identified, 25 have not appeared in any previously published 2-DE map of CSF, and 11 have not been previously reported to exist in CSF. Most of the identifications originate from an ante mortem CSF sample collected from a single hydrocephalus patient. A supplemental map created from neurologically normal patients is also presented. A webpage with protein identification and scoring information from both maps is available at http://www.leelab.org/csfmap.     

Comparative study of three immunoassays based on monoclonal antibodies for detection of the pesticide parathion-methyl in real samples

Three immunoassay systems: indirect, direct competitive enzyme-linked immunosorbent assay (IC-ELISA and DC-ELISA), fluorescence polarization immunoassay (FPIA) based on monoclonal antibodies for the detection of parathion-methyl (PM) were developed and optimized. Several PM derivatives (haptens) were conjugated to proteins and fluoresceinthiocarbamyl ethylenediamine (EDF) to obtain immunogens and competitors. The influence of immunogen and competitor structures on the assay performance was investigated. IC-ELISA was the most sensitive of all techniques developed, with a detection limit of 0.08 ng ml⁻¹, but assay time was the longest (3.5 h per 96-well microtitre plate). DC-ELISA was easier to perform and quicker (1.5 h per 96-well microtitre plate) but less sensitive than IC-ELISA (detection limit was 0.5 ng ml⁻¹). FPIA was the fastest and simplest (7 min per 10 samples) but the least sensitive (detection limit was 15 ng ml⁻¹) technique. The methods were characterized by high specificity and reproducibility. The cross-reactivity for parathion-ethyl was around 30-40% for IC-ELISA and FPIA, but significantly higher (125%) for DC-ELISA. The immunoassays were applied to the analysis of PM residues in different food and environmental matrices. Methanol extracts of vegetable, fruit and soil samples were used for the analysis. Recoveries for most spiked samples averaged between 85 and 110%. The methods developed can be used for screening of food and environmental samples for PM residues without complicated clean-up.     

Convenient synthesis of arylpropargyl aldehydes and 4-aryl-3-butyn-2-ones from arylacetylenes and amide acetals

The reaction of arylacetylenes 1 and N,N-dimethylformamide dimethylacetal (2a, DMF-DMA) afforded the corresponding arylpropargyl aldehydes 3 in moderate yields. Similarly, the reaction of 1 and N,N-dimethylacetamide dimethylacetal (2b, DMA-DMA) gave 4-aryl-3-butyn-2-ones 4.     

An efficient closed-shell singles and doubles coupled-cluster method

A reformulated set of equations for the closed-shell singles and doubles coupled-cluster (CCSD) method is presented. A computational cost of n_v⁴n₀²+7n_v³n₀³+1n_v²n₀⁴ for the n⁶ steps is obtained, where n_v is the number of virtual molecular orbitals included in the CCSD procedure, n₀ is the number of doubly occupied molecular orbitals and n=n₀+n_v. Test calculations for the cis and trans isomers of FNNF and planar and pyramidal CH₃⁻ are presented. Equilibrium structures determined with large Gaussian basis sets at the second-order Møller-Plesset (MP2) perturbation level of theory are reported and used for the other electron correlation methods. With the largest one-particle basis set (144 contracted Gaussian functions), the equilibrium geometries of cis- and trans-FNNF agree with experiment. Based on analyses of planar and pyramidal CH₃⁻ wavefunctions and the calculated inversion barrier, it is suggested that the molecular anion may not exist in a planar configuration but that autodetachment of an electron occurs before the transition state is reached. Comparisons of our new CCSD procedure demonstrate that coupled-cluster methods are not significantly more expensive than similar electron correlation techniques.