Erin M. Duffy

Prediction of drug solubility from structure.

The aqueous solubility of a drug is an important factor affecting its bioavailability. Numerous computational methods have been developed for the prediction of aqueous solubility from a compounds structure. A review is provided of the methodology and quality of results for the most useful procedures including the model implemented in the QikProp program. Viable methods now exist for predictions with less than 1 log unit uncertainty, which is adequate for prescreening synthetic candidates or design of combinatorial libraries. Further progress with predictive methods would require an experimental database of highly accurate solubilities for a large, diverse collection of drug-like molecules.

Prediction of drug solubility from Monte Carlo simulations.

Monte Carlo statistical mechanics simulations have been carried out for 150 organic solutes in water. Physically significant descriptors such as the solvent-accessible surface area, numbers of hydrogen bonds, and indices for cohesive interactions in solids are correlated with pharmacologically important properties including octanol/water partition coefficient (log P) and aqueous solubility (log S). The regression equation for log S only requires five descriptors to provide a correlation coefficient, r2, of 0.9 and rms error of 0.7 for the 150 solutes. The descriptors can form a basis for structural modifications to guide an analogues properties into desired ranges.

Crystal Structure of the Oxazolidinone Antibiotic Linezolid Bound to the 50S Ribosomal Subunit

The oxazolidinone antibacterials target the 50S subunit of prokaryotic ribosomes. To gain insight into their mechanism of action, the crystal structure of the canonical oxazolidinone, linezolid, has been determined bound to the Haloarcula marismortui 50S subunit. Linezolid binds the 50S A-site, near the catalytic center, which suggests that inhibition involves competition with incoming A-site substrates. These results provide a structural basis for the discovery of improved oxazolidinones active against emerging drug-resistant clinical strains.

Design at the atomic level: design of biaryloxazolidinones as potent orally active antibiotics.

We have developed a first generation of hybrid sparsomycin-linezolid compounds into a new family of orally bioavailable biaryloxazolidinones that have activity against both linezolid-susceptible and -resistant gram-positive bacteria as well as the fastidious gram-negative bacteria Haemophilus influenzae and Moraxella catarrahalis. The convergent synthesis of these new compounds is detailed.

Design at the atomic level : Generation of novel hybrid biaryloxazolidinones as promising new antibiotics

From the X-ray crystal structures of linezolid and the non-selective antibiotic sparsomycin, we have derived a new family of hybrid oxazolidinones. From this initial compound set we have developed a new biaryloxazolidinone scaffold that shows both potent antimicrobial activity as well as selective inhibition of ribosomal translation. The synthesis of these compounds is outlined.

Computational investigations of protein denaturation: apomyoglobin and chaotrope-arene interactions

Molecular dynamics (MD) and Monte Carlo (MC) simulations are being used to investigate protein denaturation. The calculations use the AMBER/OPLS force field with explicit representation of the solvent via the TIP3P and TIP4P models of water. The thermal denaturation of apomyoglobin has been followed in two 500 ps MD simulations at 85 °C. The resultant structures provide a detailed model of a molten globule, and close agreement is obtained with experimental data on the helical content of both native apomyoglobin and the low pH folding intermediate. The mechanism of protein denaturation by chaotropic agents is also being pursued. The possibility of direct contact between the chaotropes and aromatic sidechains is supported by MC computations of free energy profiles for the approach of urea and guanidinium ion to aromatic hydrocarbons in water.

Approaches to Protein-Ligand Binding from Computer Simulations

Accurate computation of protein-ligand binding affinities is a challenging goal with great potential value in the design of therapeutic agents. Applications of statistical mechanics simulations to the problem are considered that feature full atomic-level descriptions of the protein, ligand and aqueous environment. Basic concepts on the methodology and intermolecular interactions in solution are presented along with results of Monte Carlo simulations for binding of inhibitors by trypsin and thrombin.

Antisense lncRNA transcription drives stochastic Protocadherin α promoter choice

Stochastic and combinatorial activation of clustered Protocadherin (Pcdh) α, β, and γ gene promoters generates a cell-surface identity code in individual neurons that functions in neural circuit assembly. Here we show that Pcdhα promoter choice requires transcription of a long noncoding RNA (lncRNA) initiated from newly identified promoters located in the protein coding sequence of each Pcdhα exon. Antisense transcription of the lncRNA through the sense promoter results in its activation and in DNA demethylation of the binding sites for the CCCTC-binding protein, CTCF, located in close proximity to both sense and antisense promoters. Increased CTCF binding promotes the assembly of long-range DNA contacts between the activated promoter and a neuron-specific enhancer, thus locking in the epigenetic state of the stochastically chosen Pcdhα promoter. Examination of this hierarchical molecular mechanism in differentiating olfactory sensory neurons, suggests that antisense Pcdhα transcription is a key prerequisite for stochastic Pcdhα promoter choice in vivo.

Computational Studies of the Molecular Recognition of Halide Anions by Calix[4]Aromatics

Monte Carlo statistical mechanics simulations have been carried out to elucidate the complexation of halide anions by a bis(phenylurea) calix[4]arene and by a calix[4]pyrrole. Optimized structures for the complexes in the gas phase are found to be maintained in chlorocabon solvents, though there are reorganization penalties of 20–30 kcal/mol for the hosts to achieve the binding geometries. All complexes feature coordination of the anions through four hydrogen bonds. The binding affinities decrease with increasing size of the anion. It is demonstrated that fluoride affinities are underestimated experimentally owing to difficulties in obtaining anhydrous tetraalkylammonium fluoride salts.

Modeling Interactions with Benzene: Aryl-Aryl, Cation-π, and Chaotrope-π

The association of benzene with itself, urea, guanidinium ion, and tetramethylammonium ion has been studied in water via Monte Carlo statistical mechanics simulations. Benzene dimerization in chloroform and liquid benzene have also been considered as well as the association of urea with naphthalene in water. Calculations of potentials of mean force predict complexes to exist in each case at contact separations. The energetics and optimal structures for the gas-phase complexes have also been characterized. Differences with the solution structures are discussed along with general implications for host-guest chemistry.