Welcome! Login  

Multi-drug efflux pump: pH-dependent assembly

Wallace and Shen, in preparation, 2009.

Micelles: pKa prediction of surfactants in micelles

Wang, Wallace and Shen, in preparation, 2009.

Protein aggregation: beta-sheet assembly of amyloid fibrils

Wallace and Shen, to be submitted, 2009.

First-principles pKa prediction for proteins: the state of the art

Wallace and Shen, Methods Enzymol, under review, 2009.
Shen et al. to be submitted, 2009.

Denatured states of proteins: how to uncover specific electrostatic interactions

Shen, under review, 2009.
Shen et al, to be submitted, 2009.

Protein folding: a putative folding intermediate

Khandogin and Brooks III, J. Am. Chem. Soc. 2007. [Abstract]

The critical role of partially folded intermediate states in protein misfolding and amyloid formation has sparked interest in exploring factors that control the formation of these meta-stable species. Recent NMR experiments reported a sparsely populated intermediate of the villin headpiece domain, in which the N-terminal subdomain is unfolded under native conditions. We tested the hypothesis that the conformational state derived from a solution NMR structure represents a putative intermediate. By studying the structural and dynamic properties as well as the unique titration behavior of His41 we discovered that a hydrogen bond network serves as a linchpin for the integrity of the N-terminal subdomain.

 

Intrinsically disordered proteins: linking folding and early aggregation events in beta amyloid peptides

Khandogin and Brooks III, Proc. Natl. Acad. Sci. 2007. [Abstract]

Protein misfolding and amyloid fibril formation are often induced by a change in the environmental pH. Well-known examples include the amyloid formation of prion protein, transthyretin, beta-peptides, alpha-synuclein, which are implicated in Creutzfeldt-Jakob, amyloidoses, Alzheimer's and Parkinson's diseases, respectively. We examined the role of pH in attenuating the conformational preference and aggregation propensity of Alzheimer's beta amyloid peptides, segment (1-28) and (10-42). Based on our data, we were able to, for the first time, elucidate the atomic-level origins of the pH-modulated early aggregation step and to propose a general mechanism for the role of pH in affecting aggregation-prone states in amyloidogenic proteins.

 

Peptide folding: pH-dependent mechanism of helix-coil transitions

Khandogin, Chen and Brooks III, Proc. Natl. Acad. Sci. USA 2006. [Abstract]

Conformational equilibria of peptides containing ionizable side chains are modulated by solution pH. By revisiting an experimentally well-characterized weakly folded peptide, the C peptide from ribonuclease A, we illustrated, for the first time, the strengths and capabilities of continuous constant pH molecular dynamics in modeling pH-dependent helix-coil transitions in proteins. This work demonstrated that continuous constant pH molecular dynamics simulations can, not only give pH profiles consistent with experiment but also offer more atomic details and mechanistic information regarding the pH-dependent conformational equilibrium beyond those attainable by current experimental techniques, such as CD and NMR.

 

Protein pKa calculations: coupling protonation states sampling with conformational dynamics

Khandogin and Brooks III, Biochemistry 2006. [Abstract]

The most commonly used Poisson-Boltzmann (PB) equation based pKa prediction methods rely on a high resolution structure and ad hoc tuning of an effective protein dielectric constant. We developed a robust and efficient protocol based on short-time replica-exchange continuous constant pH molecular dynamics for accurate predictions of protein pKa's on the first-principles level. In this protocol we also incorporated an approximate function to account for salt effects as well as an improved parameterization of Generalized Born solvent model. The new approach gave a root-mean-square deviation of less than 1 pKa unit from a set of most stringent experimental benchmarks. This level of accuracy is comparable to that of the PB based methods but our method offers significant advantages. For example, we do not need high resolution structure or to adjust the protein dielectric constant. Most importantly, our method can compute protein pKa's during dynamics, folding, or other large conformational transitions.

 

Molecular dynamics simulations: development of continuous constant pH method

Khandogin and Brooks III, Biophys. J. 2005. [Abstract]

Environmental pH plays an important role in many biological processes, ranging from protein folding/misfolding, enzyme catalysis, to proton gradient driven ATP synthesis. To account for pH, traditional molecular dynamics simulations fix protonation states of protein ionizable side chains according to their standard pKa values. A novel simulation technique, continuous constant pH molecular dynamics, based on lambda-dynamics and generalized Born implicit solvent model, was developed, in which the protonation state of protein ionizable side chain is brought into equilibrium with an external bath of protons at a given pH. To further improve the physical description of ionization equilibria, we developed a two-dimensional lambda-dynamics frame work to enable multiple competing titrating sites, as is the case for histidine and carboxylate side chains. This new method brought the computed pKa's to be in quantitative agreement with experiment, predicting a correct sign for all pKa's shifts in the test proteins.

 

Continuum solvent models: development of analytic gradients for a COSMO based solvation method

Khandogin, Gregersen, Thiel, and York, J. Phys. Chem. B 2005. [Abstract]
Gregersen, Khandogin, Thiel, and York, J. Phys. Chem. B 2005. [Abstract]

Geometry optimization and transition state search in quantum calculations with conventional solvation methods present a major challenge due to the numerical instability of solvation energy and solvation energy gradients. We developed analytic gradients for a boundary element based continuum solvation model,  COSMO. This model was then implemented into a semi-empirical quantum program. Finally, we applied the semi-empirical methods with our solvation model to study the transesterification reaction of phosphodiester, a model system for the self-cleavage reaction catalyzed by the hammerhead and hairpin ribozymes.

 

Electrostatic properties, chemical reactivity and RNA binding affinity of HIV-1 nucleocapsid protein

Khandogin, Musier-Forsyth, and York, J. Mol. Biol. 2003. [Abstract]

HIV-1 nucleocapsid protein (NC) plays several important roles in the viral life-cycle and presents an attractive target for rational drug design. We characterized the chemical reactivities of thiolates toward small-molecule electrophiles, the protonation sequence of the C-terminal Zn-coordination complex, and the relative binding energies of NC to SL2 and SL3 stem-loops of the HIV-1 genomic RNA packaging signal.

 

Linear-scaling semi-empirical quantum methods: development of quantum descriptors

Khandogin and York, Proteins 2004. [Abstract]
Khandogin and York, J. Phys. Chem. B. 2002. [Abstract]
Khandogin, Hu, and York, J. Comput. Chem. 2000. [Abstract]

Divide-and-conquer based linear-scaling semi-empirical quantum methods with linear-scaling COSMO solvation model opened a door for electronic-structure calculations to be performed on biological macromolecules. Following the success of methodological development a question arose: can linear-scaling quantum methods allow us to obtain additional insights into biological macromolecules beyond what we can learn from classical molecular mechanics simulations? Towards this goal, we developed a set of quantum descriptors that characterize the electrostatic and chemical properties of proteins, DNA and RNA in solution.

 

Density functional theory: NMR spin-spin coupling constants

Khandogin and Ziegler, J. Phys. Chem. A 2000. (citation: 17) [Abstract]
Khandogin and Ziegler, Spectrochim. Acta A 1999. (citation: 25) [Abstract]

Development of Density functional theory (DFT), for which Kohn and Pople were awarded the Nobel Prize in 1998, has enabled highly accurate computational prediction of chemical properties for small to medium sized molecules. We carried out the first calculation and analysis of NMR spin-spin coupling constants in organo-metallic complexes based on DFT and perturbation theory. Later, we developed a relativistic correction scheme to predict scalar relativistic effects on nuclear spin-spin couplings of molecules containing heavy transition-metal atoms.

 

Relativistic Configuration Interaction theory: electronic states of small molecules containing heavy atoms

Khandogin et al, J. Mol. Spectrosc. 1997. [Abstract]

The two basic theories of modern physics are the theory of relativity and quantum mechanics. The relativistic Configuration Interaction (CI) based ab initio calculations on small molecules containing heavy atoms combined with spectroscopic experiments provided a frame work for testing the current level of quantum and relativistic theory in predicting the energetics and transition probabilities of electronic ground and excited states. We applied the relativistic CI method with spin-orbit coupling to calculate the potential energy surface and life time of low-lying electronic states of a tri-atomic molecule, BiOH/HBiO. The results of this theoretical investigation were later confirmed by a laser-induced fluorescence experiment.