Effects of Molecular Interactions on Miscibility and Mobility of Ibuprofen in Amorphous Solid Dispersions With Various Polymers.

Hydrogen bonds (HBs) in amorphous solid dispersions may influence physical stability through effects on both drug miscibility and mobility. Amorphous solid dispersions containing the HB-donor ibuprofen (IBP) alone or with one of four model polymers (poly(vinyl pyrrolidone) [PVP], poly(vinyl pyrrolidone/vinyl acetate) [PVP/VA], poly(vinyl acetate) [PVA], or polystyrene [PST]) were monitored by molecular dynamics simulation. HB distributions and contributions of electrostatic, van der Waals, and internal interactions to miscibility and mobility were analyzed versus drug concentration. The probability of IBP-IBP HBs decreases markedly (0.6→0.0) with dilution (100→10% drug) in PVP due to IBP-PVP HBs while dilution in the nonpolar PST has a more modest effect on IBP-IBP HB probability (0.6→0.3). Concentration-dependent Flory-Huggins interaction parameters (χ) were determined to assess drug-polymer miscibility. χIBP-PVP values were -0.9 to -1.8 with a plateau near 50% w/w PVP, whereas χIBP-PST fluctuated near zero (-0.1 to 0.3), suggesting that IBP is more soluble in PVP than in PST. χIBP-polymer values in polymers varying in pyrrolidone/acetate composition were in the order PVP (most favorable) > PVP/VA > PVA (least favorable). Decreased local mobility of IBP measured by the atomic fluctuation correlates with more IBP-PVP HBs with increasing PVP content. The opposite trend in IBP-PST may arise from IBP-IBP HB disruption on dilution.

Molecular Dynamics Simulation of Amorphous Hydroxypropylmethylcellulose and Its Mixtures With Felodipine and Water.

Understanding drug-polymer molecular interactions, their miscibility, supersaturation potential, and the effects of water uptake may be invaluable for selecting amorphous polymer dispersions that can maximize the oral bioavailability of poorly water-soluble drugs. Molecular dynamics simulations were performed using a model for hydroxypropylmethylcellulose (HPMC) resembling the substitution patterns found experimentally. HPMC at low and high water contents (0.9%-23.0% wt/wt) and mixtures with a hydrophobic drug, felodipine (FEL), were constructed. Tg values and densities after ∼30 ns aging at 298 K were close to published results. Except for hydrogen bonds (HBs) between the 5-O- and a 3-OH group in a neighboring repeat unit, HPMC oxygen atoms have a low HB probability (p < 0.1) perhaps due to shielding by surrounding substituents. Water molecules tend to be isolated at low water content while clusters were prevalent at ≥10.7% water. The Flory-Huggins FEL-HPMC interaction parameter (-0.20 ± 0.07) predicts complete miscibility at all HPMC compositions, in agreement with experiments. However, HBs between the FEL-N-H and HPMC favoring miscibility are disrupted with increasing water. Apparent diffusion coefficients versus water content were generated for water and FEL and a theory for the non-Einsteinian nature of water diffusion is proposed.

Hydrogen Bonding Interactions in Amorphous Indomethacin and Its Amorphous Solid Dispersions with Poly(vinylpyrrolidone) and Poly(vinylpyrrolidone-co-vinyl acetate) Studied Using (13)C Solid-State NMR.

Hydrogen bonding interactions in amorphous indomethacin and amorphous solid dispersions of indomethacin with poly(vinylpyrrolidone), or PVP, and poly(vinylpyrrolidone-co-vinyl acetate), or PVP/VA, were investigated quantitatively using solid-state NMR spectroscopy. Indomethacin that was (13)C isotopically labeled at the carboxylic acid carbon was used to selectively analyze the carbonyl region of the spectrum. Deconvolution of the carboxylic acid carbon peak revealed that 59% of amorphous indomethacin molecules were hydrogen bonded through carboxylic acid cyclic dimers, 15% were in disordered carboxylic acid chains, 19% were hydrogen bonded through carboxylic acid and amide interactions, and the remaining 7% were free of hydrogen bonds. The standard dimerization enthalpy and entropy of amorphous indomethacin were estimated to be -38 kJ/mol and -91 J/(mol · K), respectively, using polystyrene as the "solvent". Polymers such as PVP and PVP/VA disrupted indomethacin self-interactions and formed hydrogen bonds with the drug. The carboxylic acid dimers were almost completely disrupted with 50% (wt) of PVP or PVP/VA. The fraction of disordered carboxylic acid chains also decreased as the polymer content increased. The solid-state NMR results were compared with molecular dynamics (MD) simulations from the literature. The present work highlights the potential of (13)C solid-state NMR to detect and quantify various hydrogen bonded species in amorphous solid dispersions as well as to serve as an experimental validation of MD simulations.

Mechanistic model and analysis of doxorubicin release from liposomal formulations.

Reliable and predictive models of drug release kinetics in vitro and in vivo are still lacking for liposomal formulations. Developing robust, predictive release models requires systematic, quantitative characterization of these complex drug delivery systems with respect to the physicochemical properties governing the driving force for release. These models must also incorporate changes in release due to the dissolution media and methods employed to monitor release. This paper demonstrates the successful development and application of a mathematical mechanistic model capable of predicting doxorubicin (DXR) release kinetics from liposomal formulations resembling the FDA-approved nanoformulation DOXIL® using dynamic dialysis. The model accounts for DXR equilibria (e.g. self-association, precipitation, ionization), the change in intravesicular pH due to ammonia release, and dialysis membrane transport of DXR. The model was tested using a Box-Behnken experimental design in which release conditions including extravesicular pH, ammonia concentration in the release medium, and the dilution of the formulation (i.e. suspension concentration) were varied. Mechanistic model predictions agreed with observed DXR release up to 19h. The predictions were similar to a computer fit of the release data using an empirical model often employed for analyzing data generated from this type of experimental design. Unlike the empirical model, the mechanistic model was also able to provide reasonable predictions of release outside the tested design space. These results illustrate the usefulness of mechanistic modeling to predict drug release from liposomal formulations in vitro and its potential for future development of in vitro - in vivo correlations for complex nanoformulations.

Determination of key parameters for a mechanism-based model to predict Doxorubicin release from actively loaded liposomes.

Despite extensive study of liposomal drug formulations, reliable predictive models of release kinetics in vitro and in vivo are still lacking. Progress in the development of robust, predictive release models has been hindered by a lack of systematic, quantitative characterization of these complex drug delivery systems with respect to the myriad of factors that may influence drug release kinetics and the wide range of dissolution media/methods employed to monitor release. In this paper, the key processes and parameters needed to develop a complete mechanism-based model for doxorubicin release from actively loaded liposomal formulations resembling Doxil(®) are determined. Quantitative models must account for the driving force(s) [i.e., activity gradient(s) of the permeable species between the intraliposomal and external media] and the permeability-area product(s) for lipid bilayer transport. These factors are intertwined as membrane permeability-area products require knowledge of the drug species and concentrations that account for the release. The necessary information includes values for the drug pKa, identity of the permeable species and species permeability coefficients, a model to describe drug self-association and the relevant equilibrium constant(s), the bilayer/water partition coefficient of the predominant drug species under relevant pH conditions, and the solubility product (Ksp ) for intraliposomal precipitates that exist in such formulations.

Molecular dynamics simulation of amorphous hydroxypropyl-methylcellulose acetate succinate (HPMCAS): polymer model development, water distribution, and plasticization.

Molecular models for HPMCAS polymer have been developed for molecular dynamics (MD) simulation that attempt to mimic the complex substitution patterns in HPMCAS observed experimentally. These molecular models were utilized to create amorphous HPMCAS solids by cooling of the polymeric melts at different water contents to explore the influence of water on molecular mobility, which plays a critical role in stability and drug release from HPMCAS-based solid matrices. The densities found for the simulated amorphous HPMCAS were 1.295, 1.287, and 1.276 g/cm(3) at 0.7, 5.7, and 13.2% w/w water, indicating swelling of the polymer with increasing water content. These densities compare favorably with the experimental density of 1.285 g/cm(3) for commercial HPMCAS-(AQOAT AS-MF) supporting the present HPMCAS models as a realistic representation of amorphous HPMCAS solids. Water molecules were observed to be mostly isolated from each other at a low water content (0.7% w/w), while clusters or strands of water were pervasive and broadly distributed in size at 13.2% w/w water. The average number of first-shell water molecules (n(w)) increased from 0.17 to 3.5, though the latter is still far below that (8.9) expected for the onset of a separate water phase. Increasing water content from 0.7 to 13.2% w/w was found to reduce the T(g) by ~81 K, similar to experimental observations. Plasticization with increasing water content resulted in increasing polymer mobility and water diffusivity. From 0.7 to 13.2% w/w water, the apparent water diffusivity increased from 1.1 × 10(-9) to 7.0 × 10(-8) cm(2)/s, though non-Einsteinian behavior persisted at all water contents explored. This and the water trajectories in the polymers suggest that water diffusion at 0.7% w/w water follows a "hopping" mechanism. At a higher water content (13.2% w/w) water diffusion follows dual diffusive processes: (1) fast water motions within water clusters; and (2) slower diffusion through the more rigid polymer matrix.

Water uptake, distribution, and mobility in amorphous poly(D,L-lactide) by molecular dynamics simulation.

An explicit all-atom computational model for amorphous poly(lactide) (PLA) was developed. Molecular dynamics simulations of PLA glasses were conducted to explore various molecular interactions and predict certain physical properties. The density of a newly formed PLA glass aged for 100 ns at 298 K was 1.23 g/cm(3), close to the experimental range (1.24-1.25 g/cm(3)). The glass transition temperature (Tg = 364 K) was higher than experimental values because of the fast cooling rate (0.03 K/ps) in the simulation. The solubility parameter (20.6 MPa(1/2)) compared favorably to the literature. The water sorption isotherm obtained by relating the excess chemical potential of water in PLA to the Henry's law constant for water sorption was close to the experiment. At 0.6% (w/w), water molecules localize next to polar ester groups in PLA because of hydrogen bonding. Local mobility in PLA as characterized by the atomic fluctuation was sharply reduced near the Tg , decreasing further with aging at 298 K. The non-Einsteinian diffusion of water was found to correlate with the rotational ß-relaxation of PLA C=O groups at 298 K. A relaxation-diffusion coupling model proposed recently by the authors gave a diffusion coefficient (1.3 × 10(-8) cm(2) /s at 298 K) which is comparable to reported experimental values.

Molecular dynamics simulation of amorphous indomethacin-poly(vinylpyrrolidone) glasses: solubility and hydrogen bonding interactions.

Amorphous drug dispersions are frequently employed to enhance solubility and dissolution of poorly water-soluble drugs and thereby increase their oral bioavailability. Because these systems are metastable, phase separation of the amorphous components and subsequent drug crystallization may occur during storage. Computational methods to determine the likelihood of these events would be very valuable, if their reliability could be validated. This study investigates amorphous systems of indomethacin (IMC) in poly(vinylpyrrolidone) (PVP) and their molecular interactions by means of molecular dynamics (MD) simulations. IMC and PVP molecules were constructed using X-ray diffraction data, and force-field parameters were assigned by analogy with similar groups in Amber-ff03. Five assemblies varying in PVP and IMC composition were equilibrated in their molten states then cooled at a rate of 0.03 K/ps to generate amorphous glasses. Prolonged aging dynamic runs (100 ns) at 298 K and 1 bar were then carried out, from which solubility parameters, the Flory-Huggins interaction parameter, and associated hydrogen bonding properties were obtained. Calculated glass transition temperature (T(g)) values were higher than experimental results because of the faster cooling rates in MD simulations. Molecular mobility as characterized by atomic fluctuations was substantially reduced below the T(g) with IMC-PVP systems exhibiting lower mobilities than that found in amorphous IMC, consistent with the antiplasticizing effect of PVP. The number of IMC-IMC hydrogen bonds (HBs) formed per IMC molecule was substantially lower in IMC-PVP mixtures, particularly the fractions of IMC molecules involved in two or three HBs with other IMC molecules that may be potential precursors for crystal growth. The loss of HBs between IMC molecules in the presence of PVP was largely compensated for by the formation of IMC-PVP HBs. The difference (6.5 MPa(1/2)) between the solubility parameters in amorphous IMC (25.5 MPa(1/2)) and PVP (19.0 MPa(1/2)) suggests a small, positive free energy of mixing, although it is close to the criterion for miscibility (<7 MPa(1/2)). In contrast to the solubility-parameter method, the calculated Flory-Huggins interaction parameter (-0.61 ± 0.25), which takes into account the IMC-PVP interaction energy, predicts complete miscibility at all PVP compositions, in agreement with experimental observations. These results from MD simulations were combined with experimental values for the crystalline γ-polymorph of IMC and amorphous IMC to estimate the solubility of IMC in amorphous PVP dispersions and the theoretical enhancement in the aqueous solubility of IMC molecularly dispersed in PVP at various volume fractions.

Molecular dynamics simulation of amorphous indomethacin.

Molecular dynamics (MD) simulations have been conducted using an assembly consisting of 105 indomethacin (IMC) molecules and 12 water molecules to investigate the underlying dynamic (e.g., rotational and translational diffusivities and conformation relaxation rates) and structural properties (e.g., conformation, hydrogen-bonding distributions, and interactions of water with IMC) of amorphous IMC. These properties may be important in predicting physical stability of this metastable material. The IMC model was constructed using X-ray diffraction data with the force-field parameters mostly assigned by analogy with similar groups in Amber-ff03 and atomic charges calculated with the B3LYP/ccpVTZ30, IEFPCM, and RESP models. The assemblies were initially equilibrated in their molten state and cooled through the glass transition temperature to form amorphous solids. Constant temperature dynamic runs were then carried out above and below the T(g) (i.e., at 600 K (10 ns), 400 K (350 ns), and 298 K (240 ns)). The density (1.312 ± 0.003 g/cm(3)) of the simulated amorphous solid at 298 K was close to the experimental value (1.32 g/cm(3)) while the estimated T(g) (384 K) was ~64 degrees higher than the experimental value (320 K) due to the faster cooling rate. Due to the hindered rotation of its amide bond, IMC can exist in different diastereomeric states. Different IMC conformations were sufficiently sampled in the IMC melt or vapor, but transitions occurred rarely in the glass. The hydrogen-bonding patterns in amorphous IMC are more complex in the amorphous state than in the crystalline polymorphs. Carboxylic dimers that are dominant in α- and γ-crystals were found to occur at a much lower probability in the simulated IMC glasses while hydrogen-bonded IMC chains were more easily identified patterns in the simulated amorphous solids. To determine molecular diffusivity, a novel analytical method is proposed to deal with the non-Einsteinian behavior, in which the temporal evolution of the apparent diffusivity D(t) is described by a relaxation model such as the KWW function and extrapolated to infinite time. The diffusion coefficient found for water diffusing in amorphous indomethacin at 298 K (2.7 × 10(-9) cm(2)/s) compares favorably to results obtained in experimental IMC glasses (0.9-2.0 × 10(-9) cm(2)/s) and is mechanistically associated with ß-relaxation processes that are dominant in sub-T(g) glasses.

Quantification of trans-2,6-difluoro-4'-N,N-dimethylaminostilbene in rat plasma: application to a pharmacokinetic study.

trans-2,6-Difluoro-4'-N,N-dimethylaminostilbene (DFS), a synthetic stilbene, displayed potent pre-clinical anti-cancer activities exceeding that observed for naturally occurring resveratrol. In this study, a simple and sensitive HPLC method was developed and validated to quantify DFS in rat plasma. The lower limit of quantification (LLOQ) was 5 ng/ml. The intra- and inter-day variation in terms of relative standard deviation (RSD) was all less than 10%. The bias rate ranged from -11.5% to 6.2% while the absolute recovery ranged from 94.1 ± 2.3 to 97.3 ± 4.4%. The pharmacokinetic profiles of DFS were examined in Sprague-Dawley rats after intravenous administration (2 mg/kg). DFS displayed moderate clearance (Cl=61.5 ± 17.7 ml/min/kg) and a relatively prolonged terminal elimination half-life (t(1/2 λz)) of 351 ± 180 min. Aqueous solubility played a crucial role in the oral absorption of DFS. When DFS was given as a suspension (6 mg/kg), the absolute oral bioavailability (F) was almost negligible. However, when DFS was given in a solution prepared with hydroxypropyl-ß-cyclodextrin (6 mg/kg), the F was 12.4 ± 10.7%. Dose-escalation to 15 mg/kg resulted in much higher systemic exposure (F=40.2 ± 10.0%). As DFS is orally available after formulation with hydroxypropyl-ß-cyclodextrin and pharmacologically active systemic concentrations could be achieved after a single oral dose, the use of DFS as a cancer chemopreventive/chemotherapeutic agent is possible.

Enhanced active liposomal loading of a poorly soluble ionizable drug using supersaturated drug solutions.

Nanoparticulate drug carriers such as liposomal drug delivery systems are of considerable interest in cancer therapy because of their ability to passively accumulate in solid tumors. For liposomes to have practical utility for antitumor therapy in patients, however, optimization of drug loading, retention, and release kinetics are necessary. Active loading is the preferred method for optimizing loading of ionizable drugs in liposomes as measured by drug-to-lipid ratios, but the extremely low aqueous solubilities of many anticancer drug candidates may limit the external driving force, thus slowing liposomal uptake during active loading. This report demonstrates the advantages of maintaining drug supersaturation during active loading. A novel method was developed for creating and maintaining supersaturation of a poorly soluble camptothecin analogue, AR-67 (7-t-butyldimethylsilyl-10-hydroxycamptothecin), using a low concentration of a cyclodextrin (sulfobutylether-ß-cyclodextrin) to inhibit crystallization over a 48 h period. Active loading into liposomes containing high concentrations of entrapped sodium or calcium acetate was monitored using drug solutions at varying degrees of supersaturation. Liposomal uptake rates increased linearly with the degree of supersaturation of drug in the external loading solution. A mathematical model was developed to predict the rate and extent of drug loading versus time, taking into account the chemical equilibria inside and outside of the vesicles and the transport kinetics of various permeable species across the lipid bilayer and the dialysis membrane. Intraliposomal sink conditions were maintained by the high internal pH caused by the efflux of acetic acid and exchange with AR-67, which undergoes lactone ring-opening, ionization, and membrane binding in the interior of the vesicles. The highest drug to lipid ratio achieved was 0.17 from a supersaturated solution at a total drug concentration of 0.6 mg/ml. The rate and extent of loading was similar when a different intraliposomal metal cation (sodium) was used instead of calcium. The proposed method may have general application in overcoming the formulation challenges associated with the liposomal delivery of poorly soluble, ionizable anticancer agents.

Pharmacokinetic modeling to assess factors affecting the oral bioavailability of the lactone and carboxylate forms of the lipophilic camptothecin analogue AR-67 in rats.

PURPOSE: Camptothecin analogues are anticancer drugs effective when dosed in protracted schedules. Such treatment is best suited for oral formulations. AR-67 is a novel lipophilic analogue with potent efficacy in preclinical models. Here we assessed factors that may influence its oral bioavailability in rats. METHODS: Plasma pharmacokinetic (PK) studies were conducted following administration of AR-67 lactone or carboxylate doses alone or after pre-dosing with inhibitors of the efflux transporters P-gp and Bcrp. A population PK model that simultaneously fitted to oral and intravenous data was used to estimate the bioavailability (F) and clearance of AR-67. RESULTS: An inverse Gaussian function was used as the oral input into the model and provided the best fits. Covariate analysis showed that the bioavailability of the lactone, but not its clearance, was dose dependent. Consistent with this observation, the bioavailability of AR-67 increased when animals were pretreated orally with GF120918 or Zosuquidar. CONCLUSION: Absorption of AR-67 is likely affected by solubility of its lactone form and interaction with efflux pumps in the gut. AR-67 appears to be absorbed as the lactone form, most likely due to gastric pH favoring its formation and predominance. F increased at higher doses suggesting saturation of efflux mechanisms.

Factors affecting the in vivo lactone stability and systemic clearance of the lipophilic camptothecin analogue AR-67.

PURPOSE: The narrow efficacy-toxicity window of anticancer agents necessitates understanding of factors contributing to their disposition. This is especially true for camptothecins as they exist in the lactone and carboxylate forms with each moiety differentially interacting with efflux or uptake transporters. Here we determined the disposition of the lactone and carboxylate forms of AR-67, a 3(rd) generation camptothecin analogue. METHODS: Pharmacokinetic studies were conducted in rats given intravenous boluses of AR-67 lactone or carboxylate with or without pharmacologic inhibitor pretreatment (GF120918 or rifampin). Pharmacokinetic modeling was used to estimate clearances, while simulations assessed the impact of clearance changes on overall AR-67 exposure. RESULTS: Our modeling showed that carboxylate clearance was 3.5-fold higher than that of the lactone. GF120918 decreased lactone clearance only, but rifampin decreased both lactone and carboxylate clearances. Simulations showed that decreasing carboxylate clearance, which controls the overall drug disposition, leads to significant increase in AR-67 exposure. CONCLUSION: The apparent in vivo blood stability of AR-67 is partly dependent on the increased carboxylate clearance. This may have clinical implications for populations with single nucleotide polymorphisms that impair the function of uptake transporter genes (e.g., SLCO1B1), which are potentially responsible for AR-67 carboxylate clearance.

Development of structure-lipid bilayer permeability relationships for peptide-like small organic molecules.

Computational methods to estimate passive membrane permeability coefficients of organic molecules, including peptides, would be valuable in understanding various biological processes associated with molecular transport across cell membranes and in reducing the time required for screening developability properties of new drug candidates. This study explores the suitability of fragment-based linear free energy relationships (LFERs) to predict lipid bilayer permeability coefficients and decadiene/water partition coefficients of a set of 47 model permeants. The inclusion of mono-, di-, and tripeptides comprised of glycine, alanine, and sarcosine residues in the database presented added challenges due to the apparent lack of independence of the contribution of the backbone amide residue in peptides to the free energy of transfer (Delta(Delta G degrees ) -CONH-) from water to organic solvents or to the bilayer barrier domain. In order to elucidate the impact of neighboring group effects on Delta(Delta G degrees ) -CONH-, a series of RGZ glycine (G)-containing peptides having an additional -NHCH 2CO- residue compared to their RZ counterparts were synthesized, where R = acetyl (Ac-), 4-carboxymethylphenyl acetyl (CMPA-), or 4-methylphenyl acetyl (MPA-), and Z = -OH, -OMe, -NHMe, or -NMe 2. While variations in R had no significant impact on Delta(Delta G degrees ) -Gly-, significant effects of neighboring ( i + 1) Z substituents at the C-terminus were revealed both in studies of the relative transport of RGZ/RZ compound pairs across DOPC bilayers and partitioning between water and 1,9-decadiene (a bulk solvent with a similar chemical selectivity to the barrier domain of DOPC/eggPC bilayers). The proximity effects decline when the bulk solvent used in partitioning studies is 1-octanol, suggesting a possible role for intramolecular hydrogen bonding in the observed nonadditivity of Delta(Delta G degrees ) -CONH-. A new LFER for predicting decadiene/water partition coefficients was developed by including the contributions of polar fragments, total nonpolar surface area of nonpolar fragments, and correction factors to account for the effects of i + 1 substituents in peptides on the group contribution of the peptide backbone amide bond to the free energy of transfer. This LFER could be used to predict lipid bilayer permeability coefficients by including an additional term to account for the added influence of molecular size on bilayer permeability.

Liposome transport of hydrophobic drugs: gel phase lipid bilayer permeability and partitioning of the lactone form of a hydrophobic camptothecin, DB-67.

The design of liposomal delivery systems for hydrophobic drug molecules having improved encapsulation efficiency and enhanced drug retention would be highly desirable. Unfortunately, the poor aqueous solubility and high membrane binding affinity of hydrophobic drugs necessitates extensive validation of experimental methods to determine both liposome loading and permeability and thus the development of a quantitative understanding of the factors governing the encapsulation and retention/release of such compounds has been slow. This report describes an efflux transport method using dynamic dialysis to study the liposomal membrane permeability of hydrophobic compounds. A mathematical model has been developed to calculate liposomal membrane permeability coefficients of hydrophobic compounds from dynamic dialysis experiments and partitioning experiments using equilibrium dialysis. Also reported is a simple method to study the release kinetics of liposome encapsulated camptothecin lactone in plasma by comparing the hydrolysis kinetics of liposome entrapped versus free drug. DB-67, a novel hydrophobic camptothecin analogue has been used as a model permeant to validate these methods. Theoretical estimates of DB-67 permeability obtained from the bulk solubility diffusion model and the "barrier-domain" solubility diffusion model are compared to the experimentally observed value. The use of dynamic dialysis in drug release studies of liposome and other nanoparticle formulations is further discussed and experimental artifacts that can arise without adequate validation are illustrated through simulations.

Liposomal drug transport: a molecular perspective from molecular dynamics simulations in lipid bilayers.

Computational methods to predict drug permeability across biomembranes prior to synthesis are increasingly desirable to minimize the investment in drug design and development. Significant progress in molecular dynamics (MD) simulation methodologies applied to lipid bilayer membranes, for example, is making it possible to move beyond characterization of the membranes themselves to explore various thermodynamic and kinetic processes governing membrane binding and transport. Such methods are also likely to be directly applicable to the design and optimization of liposomal delivery systems. MD simulations are particularly valuable in addressing issues that are difficult to explore in laboratory experiments due to the heterogeneity of lipid bilayer membranes at the molecular level. Insights emerging from MD simulations are contributing to an understanding of which regions within bilayers are most and least favored by solutes at equilibrium as the solute structure is varied, local diffusivities of permeants, and the origin of the amplified selectivity to permeant size imposed by lipid bilayer membranes, particularly as changes in composition increase acyl chain ordering.

Molecular dynamics simulations and experimental studies of binding and mobility of 7-tert-butyldimethylsilyl-10-hydroxycamptothecin and its 20(S)-4-aminobutyrate ester in DMPC membranes.

The enhanced permeability and retention of liposomes in solid tumors makes liposomal formulations attractive for the targeting of various antitumor agents. This study explores the binding, orientation, and dynamic properties of a potent topoisomerase I inhibitor, 7-tert-butyldimethylsilyl-10-hydroxycamptothecin (DB-67), and its 20(S)-4-aminobutyrate ester prodrug (DB-67-AB) in DMPC liposomes by molecular dynamics (MD) simulations and experimental studies. MD simulations of an all-atom and fully hydrated liquid-crystalline bilayer (2 x 36 DMPC lipids) containing single molecules of DB-67 and DB-67-AB were conducted for up to 50 ns. Membrane/water partition coefficients for DB-67 and DB-67-AB vs pH were determined by ultracentrifugation. Fluorescence spectra and/or steady-state anisotropies were measured in various solvents and in DMPC liposomes. Kinetics for the reversible DB-67 lactone ring-opening in the presence and absence of DMPC liposomes were determined by HPLC with fluorescence detection. During the entire simulation time both DB-67 and DB-67-AB were located on the bilayer membrane near the polar ester groups of DMPC. The average depth of penetration for DB-67 and DB-67-AB was similar (12.4-13.2 A) with the prodrug's protonated amino group strongly solvated by surface water and lipid phosphate groups. Binding and fluorescence experiments revealed only a modest reduction in the binding affinity upon attachment of the ionized 4-aminobutyrate group onto DB-67. The binding microenvironment polarity resembles that of a polar solvent such as EtOH and DMSO. Kinetics experiments confirmed that DB-67 lactone hydrolysis is inhibited in the presence of DMPC liposomes, consistent with the reduced exposure of its lactone ring to water, as observed in the simulations. Both bound DB-67 and bound DB-67-AB have nonrandom orientations and reduced mobility in the membrane, especially for diffusion normal to the bilayer surface, and rotational relaxation, both of which are > or =2 orders of magnitude slower than in bulk water. MD simulations correctly predicted the high binding affinities for DB-67-AB to DMPC bilayers, protection of bound DB-67 toward lactone hydrolysis, and the lack of a substantial reduction in binding for the 20(S)-4-aminobutyrate prodrug of DB-67.

Conformational structure, dynamics, and solvation energies of small alanine peptides in water and carbon tetrachloride.

The rate-limiting barrier for peptide transport across lipid bilayers is the nonpolar hydrocarbon interior. Permeating peptides may undergo conformational changes during their transfer from an aqueous solution into the barrier domain, thus facilitating peptide transport. To test this hypothesis, all-atom and explicit-solvent molecular dynamics (MD) simulations have been conducted on a series of small peptides, p-toluyl-Ala(n) (n = 0-3) used previously in transport experiments, to explore their conformational structures, dynamics and solvation free energies in water and carbon tetrachloride (CCl(4)). The conformations of the p-toluyl alanine di- and tri-peptides in water were found to be far from random coils, with P(II) and alpha(R) dominating but with smaller populations of seven-membered (c(7)) and five-membered rings (c(5)). In contrast, the seven-membered ring, c(7), along with c(5) dominated in CCl(4). These results indicate that the conformational preferences of the alanine peptides are highly sensitive to solvent. Dynamically, stable seven-membered ring formation occurred on a time scale of 10 ps while larger ring-sizes (e.g., 10-membered rings) were observed much less frequently. The values of adjacent torsional angles (phi(1), psi(1)) were dependent on neighboring torsional angles. Thermal motions of neighboring torsions leading to transitions between c(7), c(5), alpha(R), and P(II) conformers were highly cooperative while longer range correlations between transitions of adjacent sets of torsions (phi(1), psi(1)) and (phi(2), psi(2)) were less evident. Peptide folding in CCl(4) lowers the intramolecular electrostatic energies. This, along with hydrophobic interactions, favors partitioning into CCl(4). These effects only partially offset other types of intramolecular interactions and peptide-solvent polar interactions that are more favorable in water, leading to net transfer free energies (3-7 kcal/mol) that disfavor peptide transfer from water into carbon tetrachloride.

Distribution and effect of water content on molecular mobility in poly(vinylpyrrolidone) glasses: a molecular dynamics simulation.

PURPOSE: This work explores the distribution of water and its effects on molecular mobilities in poly(vinylpyrrolidone) (PVP) glasses using molecular dynamics (MD) simulation technology. METHODS: PVP glasses containing 0.5% and 10% w/w water and a small amount of ammonia and Phe-Asn-Gly were generated. Physical aging processes and associated structural and dynamic properties were monitored vs. time for periods up to 0.1 micros by MD simulation. RESULTS: Increasing water content from 0.5% to 10% w/w was found to reduce the Tg by about 90 K and increase the rates of volume and enthalpy relaxation. At 0.5% w/w, water molecules are mostly isolated and uniformly distributed while at 10% w/w, water distribution is markedly heterogeneous, with strands of water molecules occupying channels between the polymer chains. At 10% w/w, each water molecule has an average of 2.0 neighboring water molecules. The plasticization effects of water were revealed in diffusion coefficient increases of 3.7-, 7.3-, and 7.6-fold for water, ammonia, and the individual polyvinylpyrrolidone segments, respectively, and in shorter relaxation times (37- to 47-fold) for rotation of polymer segments with an elevation in water content from 0.5% to 10% w/w. Water diffusivity was found to linearly correlate with the number of neighboring water molecules. Rotation of the PVP segments is comprised of a fast wobble motion within a highly restrained cavity and a slow rotation over a wider angular space. Only the slow rotation was shown to be significantly affected by water content. CONCLUSIONS: Water distribution in the PVP glass is highly heterogeneous at 10% w/w water, reflecting the formation of water strands or small clusters rather than complete phase separation. Local enhancement of mobility with increasing water content has been demonstrated using MD simulations.

A molecular dynamics simulation of reactant mobility in an amorphous formulation of a peptide in poly(vinylpyrrolidone).

The reaction pathways available for chemical decomposition in amorphous solids are determined in part by the relative mobilities of the potential reactants. In this study, molecular dynamics simulations of amorphous glasses of polyvinylpyrrolidone (PVP) containing small amounts of water, ammonia, and a small peptide, Phe-Asn-Gly, have been performed over periods of up to 100 ns to monitor the aging processes and associated structural and dynamic properties of the PVP segments and embedded solutes. Glass transition temperatures, Tg, were detected by changes in slopes of the volume-temperature profiles and the internal energy-temperature profiles for the inherent structures upon cooling at different rates. Analyses of the molecular trajectories below Tg reveal both temporal and spatial heterogeneity in polymer and solute mobility, with each molecule or part of a molecule displaying quite different relaxation behaviors for translational, rotational, and/or conformational motions. Rotations of individual polymer segments on the time scale up to 100 ns, though far from complete, are described by the Kohlrausch-Williams-Watts stretched exponential function with relaxation times tau on the order of 10-2.8 x 10(4) micros at an averaged stretching parameter beta of 0.39. The rotation rates are, on the average, faster for the side chains and for segments near the ends of the chains than for the backbones and segments near the middle of the chains. In contrast to their behavior in water, solute diffusive motions in the glassy polymer exhibit non-Einsteinian behavior over the time scale of the simulations characterized by two types of motion: (1) entrapments within relatively fluid microdomains surrounded by a matrix of relatively immobile polymer chains; and (2) jumps between microdomains with greater probability of hopping back to the solute's previous location. The average jump length and frequency are highly dependent on solute size, being much smaller for the tripeptide, Phe-Asn-Gly, than for water and ammonia. The diffusivities of water and ammonia, solutes capable of forming hydrogen bonds with the lactam residues within the polymer segments, are significantly reduced by strong electrostatic interactions. The conformational preferences of Phe-Asn-Gly were compared in the amorphous polymer and water to detect differences in the degree to which the tripeptide may be predisposed toward deamidation of the asparagine side chain in these environments. Although only minor differences are evident in peptide conformation, the conformational dynamics for the peptide embedded in the glassy polymer are characterized by a higher energy barrier between conformational states and 2.5-44-fold larger relaxation times for the dihedral angles of interest than in water. However, in the context of peptide deamidation, these differences may be of secondary importance in comparison to the more than two to three orders of magnitude reduction in the diffusivities of water, ammonia, and the tripeptide in PVP.