On modeling biomolecular-surface nonbonded interactions: application to nucleobase adsorption on single-wall carbon nanotube surfaces

In this work we explored the selectivity of single nucleobases towards adsorption on chiral single-wall carbon nanotubes (SWCNTs) by density functional theory calculations. Specifically, the adsorption of molecular models of guanine (G), adenine (A), thymine (T), and cytosine (C), as well as of AT and GC Watson-Crick (WC) base pairs on chiral SWCNT C(6, 5), C(9, 1) and C(8, 3) model structures, was analyzed in detail. The importance of correcting the exchange-correlation functional for London dispersion was clearly demonstrated, yet limitations in modeling such interactions by considering the SWCNT as a molecular model may mask subtle effects in a molecular-macroscopic material system. The trend in the calculated adsorption energies of the nucleobases on same diameter C(6, 5) and C(9, 1) SWCNT surfaces, i.e., G > A > T > C, was consistent with related computations and experimental work on graphitic surfaces, however contradicting experimental data on the adsorption of single-strand short homo-oligonucleotides on SWCNTs that demonstrated a trend of G > C > A > T (Albertorio et al 2009 Nanotechnology 20 395101). A possible role of electrostatic interactions in this case was partially captured by applying the effective fragment potential method, emphasizing that the interplay of the various contributions in modeling nonbonded interactions is complicated by theoretical limitations. Finally, because the calculated adsorption energies for Watson-Crick base pairs have shown little effect upon adsorption of the base pair farther from the surface, the results on SWCNT sorting by salmon genomic DNA could be indicative of partial unfolding of the double helix upon adsorption on the SWCNT surface.

On understanding proton transfer to the biocatalytic [Fe-Fe](H) sub-cluster in [Fe-Fe]H(2)ases: QM/MM MD simulations

Proton transfer to the [Fe-Fe](H) sub-cluster in the Desulfovibrio desulfuricans (DdH) and Clostridium pasteurianum (CpI) [Fe-Fe] hydrogenases was investigated by a combination of first principles and empirical molecular dynamics simulations. Pathways that can be inferred from the X-ray crystal structures of DdH and CpI, i.e., (Glu159→Ser198→Glu156→water460→Cys178→DTMA([Fe-Fe](H)) and (Glu282→Ser319→Glu279→water612→Cys299), respectively, were considered. Proton transfer from Cys178 to DTMA in the [Fe-Fe](H) sub-cluster in DdH was readily observed in our results, specifically when [Fe-Fe](H) was in the reduced state ([Fe(I)-Fe(I)]) or in the mixed valence state for the protonated distal iron Fe(d) ([Fe(I)-Fe(II)-H(-)](H)). A concerted mechanism is proposed, where proton transfer in DdH from Glu159 to Glu156 via Ser198 and Glu156 to Cys178 via water460 readily occurred, as well as from Glu282 to Glu279 via Ser319 and Glu279 to Cys299 via water612 in CpI. The theoretical prediction of the proton transfer characteristics is consistent with the assumed biocatalytic mechanism of the [Fe-Fe] hydrogenases in which the proton binds at Fe(d), providing confirmation that has not been explored so far. The computational results were qualitatively validated by the agreement with experimental hydrogen production activity data for mutated CpI enzymes, relative to the wild-type protein. Finally, the insight provided by the simulations, combined, in part, with experimental validation, are important for establishing an approach in future exploration of proton transfer to the active site in this class of enzymes, and possibly also for biomimetic analogs.

Understanding effects of molecular adsorption at a single-wall boron nitride nanotube interface from density functional theory calculations

In this paper, we explored computationally the feasibility of modulating the bandgap in a single-wall BN nanotube (BNNT) upon noncovalent adsorption of organic molecules, combined with the application of a transverse electric field. Effects of analytes' physisorption on the surface of BNNTs regarding structural and electronic properties were delineated. Relatively large binding energies were calculated, however, with minimal perturbation of the structural framework. Electronic structure calculations indicated that the bandgap of BNNTs can be modified by weak adsorption due to the presence of adsorbate states in the gap of the host system. Furthermore, we have shown that the application of a transverse electric field can tune the bandgap by shifting adsorbate states, consistent with calculated current-voltage characteristics.

Quantum descriptors for predictive toxicology of halogenated aliphatic hydrocarbons

In order to improve Quantitative Structure-Activity Relationships (QSARs) for halogenated aliphatics (HA) and to better understand the biophysical mechanism of toxic response to these ubiquitous chemicals, we employ improved quantum-mechanical descriptors to account for HA electrophilicity. We demonstrate that, unlike the lowest unoccupied molecular orbital energy, ELUMO, which was previously used as a descriptor, the electron affinity can be systematically improved by application of higher levels of theory. We also show that employing the reciprocal of ELUMO, which is more consistent with frontier molecular orbital (FMO) theory, improves the correlations with in vitro toxicity data. We offer explanations based on FMO theory for a result from our previous work, in which the LUMO energies of HA anions correlated surprisingly well with in vitro toxicity data. Additional descriptors are also suggested and interpreted in terms of the accepted biophysical mechanism of toxic response to HAs and new QSARs are derived for various chemical categories that compose the data set employed. These alternate descriptors provide important insight and could benefit other classes of compounds where the biophysical mechanism of toxic response involves dissociative attachment.

Improved QSARs for predictive toxicology of halogenated hydrocarbons

In our continuing efforts to provide a predictive toxicology capability, we seek to improve QSARs (quantitative structure-activity relationships) for chemicals of interest. Currently, although semi-empirical molecular orbital methods are hardly the state of the art for studying small molecules, AM1 calculations appear to be the method of choice when calculating quantum-chemical descriptors. However, with the advent of modern computational capabilities and the development of fast algorithms, ab initio molecular orbital and first principles density functional methods can be expeditiously applied in current QSAR studies. We present a study on halogenated alkanes to assess whether more accurate quantum methods result in QSARs that correlate better with experimental data. Furthermore, improved QSARs can also be obtained through development of new descriptors with explicit physical interpretations that should lead to better understanding of the mechanisms involved in the toxic response. We show that descriptors calculated from chemical intermediates may be useful in future QSARs.

Molecular dynamics of bacteriorhodopsin

A model of bacteriorhodopsin (bR), with a retinal chromophore attached, has been derived for a molecular dynamics simulation. A method for determining atomic coordinates of several ill-defined strands was developed using a structure prediction algorithm based on a sequential Kalman filter technique. The completed structure was minimized using the GROMOS force field. The structure was then heated to 293 K and run for 500 ps at constant temperature. A comparison with the energy-minimized structure showed a slow increase in the all-atom RMS deviation over the first 200 ps, leveling off to approximately 2.4 A relative to the starting structure. The final structure yielded a backbone-atom RMS deviation from the crystallographic structure of 2.8 A. The residue neighbors of the chromophore atoms were followed as a function of time. The set of persistent near-residue neighbors supports the theory that differences in pKa values control access to the Schiff base proton, rather than formation of a counterion complex.

Biomolecular structure prediction at a low resolution using a neural network and the double-iterated Kalman filter technique

We report the application of an integrated computational approach for biomolecular structure determination at a low resolution. In particular, a neural network is trained to predict the spatial proximity of C-alpha atoms that are less than a given threshold apart, whereas a Kalman filter algorithm is employed to outline the biomolecular fold, with a constraints set that includes these pairwise atomic distances, and the distances and angles that define the structure as it is known from the protein's sequence. The results for Crambin demonstrate that this integrated approach is useful for molecular structure prediction at a low resolution and may also complement existing experimental distance data for a protein structure determination.

On the three-dimensional structure of the Hydra head activator neuropeptide

The structure of the Hydra head activator (HHA) neuropeptide has been previously studied using NMR, CD, and Raman spectroscopy and determined to contain 62-67% anti-parallel beta-pleated sheet, and predicted to assume a beta-turn near the amino terminus. We have utilized spectroscopic data with the Double-Iterated Kalman Filter (DIFK) technique and CHARMm molecular mechanics to produce a molecular model of the HHA neuropeptide. Consistent with the secondary structure prediction, an anti-parallel beta-pleated sheet topology was evident from the serine amino acid to the carboxyl terminus. Additionally, a beta-turn occurred near the amino carboxyl terminus. Results indicate that fluctuations occurring at both termini may serve to stabilize the structure ultimately allowing the amino terminus access to its receptor protein.

The solution structures of Escherichia coli trp repressor and trp aporepressor at an intermediate resolution

We have determined the solution structures and examined the dynamics of the Escherichia coli trp repressor (a 25-kDa dimer), with and without the co-repressor L-tryptophan, from NMR data. This is the largest protein structure thus far determined by NMR. To obtain a set of data sufficient for a structure determination it was essential to resort to isotopic spectral editing. Line broadening observed in this molecular mass range precludes for the most part the measurement of coupling constants and stereospecific assignments, with the inevitable result that the attainable resolution of the final structure will be somewhat lower than the resolution reported for smaller proteins and peptides. Nevertheless the general topology of the protein can be deduced from the subsets of NOEs defining the secondary and tertiary structure, providing a basis for further refinement using the full set of NOEs and energy minimization. We report here (a) an intermediate resolution structure that can be deduced from NMR data, covalent, angular and van-der-Waals constraints only, without resort to detailed energy calculations, and (b) the limits of uncertainty within which this structure is valid. An examination of these structures combined with backbone amide exchange data shows that even at this resolution three important conclusions can be drawn: (a) the protein structure changes upon binding tryptophan; (b) the putative DNA binding region is much more flexible than the core of the molecule, with backbone amide proton exchange rates 1000 times faster than in the core; (c) the binding of tryptophan stabilizes the repressor molecule, which is reflected in both the appearance of additional NOEs, and in the slowing of backbone proton exchange rates by factors of 3-10. Sequence-specific 1H-NMR assignments and the secondary structure of the holopressor (L-tryptophan-bound form) have been reported previously [C. H. Arrowsmith, R. Pachter, R. B. Altman, S. B. Iyer & O. Jardetzky (1990) Biochemistry 29, 6332-6341]. Those for the trp aporepressor (L-tryptophan-free form), made using the same methods and conditions as described in the cited paper, are reported here. The secondary structure of the aporepressor was calculated from sequential and medium-range NOEs and is the same as reported for the holorepressor except that helix E is shorter. The tertiary solution structures for both forms of the repressor were calculated from long-range NOE data.(ABSTRACT TRUNCATED AT 400 WORDS)

Sequence-specific 1H NMR assignments and secondary structure in solution of Escherichia coli trp repressor

Sequence-specific 1H NMR assignments are reported for the active L-tryptophan-bound form of Escherichia coli trp repressor. The repressor is a symmetric dimer of 107 residues per monomer; thus at 25 kDa, this is the largest protein for which such detailed sequence-specific assignments have been made. At this molecular mass the broad line widths of the NMR resonances preclude the use of assignment methods based on 1H-1H scalar coupling. Our assignment strategy centers on two-dimensional nuclear Overhauser spectroscopy (NOESY) of a series of selectively deuterated repressor analogues. A new methodology was developed for analysis of the spectra on the basis of the effects of selective deuteration on cross-peak intensities in the NOESY spectra. A total of 90% of the backbone amide protons have been assigned, and 70% of the alpha and side-chain proton resonances are assigned. The local secondary structure was calculated from sequential and medium-range backbone NOEs with the double-iterated Kalman filter method [Altman, R. B., & Jardetzky, O. (1989) Methods Enzymol. 177, 218-246]. The secondary structure agrees with that of the crystal structure [Schevitz, R., Otwinowski, Z., Joachimiak, A., Lawson, C. L., & Sigler, P. B. (1985) Nature 317, 782], except that the solution state is somewhat more disordered in the DNA binding region and in the N-terminal region of the first alpha-helix. Since the repressor is a symmetric dimer, long-range intersubunit NOEs were distinguished from intrasubunit interactions by formation of heterodimers between two appropriate selectively deuterated proteins and comparison of the resulting NOESY spectrum with that of each selectively deuterated homodimer. Thus, from spectra of three heterodimers, long-range NOEs between eight pairs of residues were identified as intersubunit NOEs, and two additional long-range intrasubunits NOEs were assigned.

Quantum-chemical studies of aflatoxin B1, sterigmatocystin and versicolorin A, and a comparison with their mutagenic activity

The INDO atomic charges q and Wiberg bond indices p (in electrons) were calculated for aflatoxin B1, sterigmatocystin and versicolorin A. The C-2-C-3 bond in these compounds has the same bond order and is predicted to be the most reactive towards epoxidation. The electronic effects do not explain the observed differences in mutagenicity and toxicity.