Executive Summary

Science Driver 3: Biomolecular MaterialsThe goals of the Science Driver 3 (SD3) team are to develop, apply, and experimentally validate multi-scale computational tools to enable the design of novel unimolecular and self-assembled drug delivery vehicles. The barriers to achieving these goals are the complexity associated with modeling systems with molecular detail from nm to mm in size and up to ms or ms in time required to make meaningful predictions. SD3 is composed of multi-disciplinary teams of scientists at five institutions (Tulane, LSU, LSU-Ag, UNO, and LA Tech), combining theoretical/computational scientists with experimentalists to build predictive targeted drug delivery models. Over the first year, a solid foundation was laid recruiting graduate and post-doctoral scholars who developed models and performed preliminary studies to map out interesting systems to explore. In year two, studies have advanced to examine delivery vehicle shapes, conformations, and free energies in depth, as well as to synthesize new vehicles to determine delivery efficacy.

B.1.3 SD3: Biomolecular Materials: [H. Ashbaugh (Tulane), D. Moldovan (LSU)]SD3 encompasses multi-disciplinary teams of scientists, post-docs, and graduate students from five institutions (Tulane, LSU, LSU-Ag, UNO, and LATech) that combine theoretical/computational scientists with experimentalists working together to develop methodologies for building new predictive models of unimolecular and self-assembled vehicles for targeted drug delivery. The SD3 research projects are conducted under two focus areas: focus 1: Unimolecular delivery vehicles, and focus 2: Self-assembled delivery vehicles.

Methods Employed/DevelopedNew computational methods are being developed on several fronts within SD3. Specifically, Ashbaugh and Rick are working to port/validate the previously developed Replica Exchange with Dynamical Sampling1 (REDS) for sampling equilibrium biomolecule conformations in solvent onto LONI. Moldovan and Nikitopoulos are continuing to work with LAMMPS developers at Sandia to link explicit molecular dynamics simulations with commercial CFD solving packages. Mobley has developed a new automated tool for performing high throughput free energy analyses on thousands of systems. Applications of these methods are described below under research accomplishments

Goals and MilestonesThe goals of the SD3 team are to develop, apply, and experimentally validate multi-scale computational tools to enable the design of novel unimolecular and self-assembled drug delivery vehicles. To achieve these goals the following milestones were set for the second year: i) Synthesize modular library of core molecules and amphiphilic side chains to explore encapsulation based on architecture and chemistry, ii) Develop new inter-atomic interaction potentials and new coarse-grained force fields for systems containing both biological and non-biological molecules; iii) Develop new hybrid MD/continuum and coarse-grained and accelerated simulation strategies to link length and time scales in biological systems; iv) Synthesize, characterize, and assess new trans-membrane drug delivery systems.

Recruiting, Hiring, and Project CoordinationIn the first year of this project SD3 successfully filled the majority of the graduate student and post-doctoral positions allocated to this driver. In year two, we filled vacant positions, including a new post-doctoral scholar, S. Paraeswaran, and a new graduate student, Janene Baker, both working under S. Rick and D. Mobley at UNO and in collaboration with H. Ashbaugh at Tulane. Additionally in year two, Anne Robinson, was recruited as department chair of Chemical and Biomolecular Engineering at Tulane and joined LA-SiGMA as a faculty member of SD3.

Research AccomplishmentsUtilizing the resources made available through LA-SiGMA over year two we continue to make significant progress in achieving the research goals and milestones of both SD3 focus areas. Below we detail progress in each research project undertaken over the second funding year.

Focus 1: Unimolecular delivery vehiclesSimulation of polymeric drug delivery carriers: [H. Ashbaugh (Tulane), L. Liu (Tulane), S. Grayson (Tulane), S. Rick (UNO), and S. Parameswaran (UNO)]. In the Ashbaugh lab a study of the temperature induced collapse transition of poly(n-isopropylacrylamide) in aqueous solution was performed as a potential trigger for drug release from gels. Large-scale simulations were carried out at LONI using replica exchange molecular dynamics, allowing an unprecedented characterization of conformational populations over a broad temperature range. As a result, the thermodynamics of the collapse transition were characterized for the first time via a two-state model, demonstrating the hydrophobic origin of this unusual behavior. This project advances milestones ii and iii.

The Rick and Ashbaugh groups are presently developing models and strategies for simulating the linear and cyclic amphiphilic polymer delivery vehicles synthesized by Grayson.2

DFT calculations have been carried out to assign partial charges and preliminary simulations have been performed to examine vehicle conformations in water (Figure A). Over the next year, simulations will be carried out using the REDS1 algorithm developed by Rick to more comprehensively sample polymer conformations in polar and non-polar environments. This project advances milestone iii.

Synthesis of new unimolecular drug delivery vehicles: [S. Grayson (Tulane) and Y. Yang (Tulane)]. Research in the Grayson laboratory has shifted towards efficient means of assembling novel amphiphilic polymeric components via efficient “click” couplings.3 The two conjugation reactions that exhibit the most utility are the copper-catalyzed azide-alkyne coupling, and the thiol-ene reaction. The versatility of this modular approach – the synthesis of individual amphiphilic components and the assembly of multifunctional amphiphilic polymers – requires substantial input from modeling in order to direct synthetic efforts. The modeling efforts of Ashbaugh and Rick (above) provide insight into the most effective means for preparing amphiphilic systems which are likely to assemble into discreet nanometer sized carriers that are of most use as biological delivery vehicles. This project advances milestones i and iv.

Nanoparticle delivery to cancer tissues: [P. DeRosa (LATech)]. Diffusion of nanoparticles in blood vessels and tumors was simulated using a combination of fluid dynamics and Monte Carlo. The model accurately predicts particle delivery to tumors as a function of relevant parameters such us blood pressure, particle size, pore size, interstitial pressure and nanoparticle concentration. This year activities were aligned with Patrick O’Neal’s (LATech) experimental group, and common ground for validating our simulation predictions have been established. This project advances milestones iii and iv.

Focus 2: Self-Assembled delivery vehiclesSimulation studies of span-80 assembly [D. Moldovan (LSU), C. Sabliov (LSU-Ag), D. Nikitopoulos (LSU), H. Ashbaugh (Tulane), B. Thakur (LSU), B. Novak (LSU), R. Kumuditha (LSU), J. Lin (LSU), and K. Xia (LSU)]: Span-80 is nonionic surfactant whose assembled structures might be used for drug delivery. Improved potential energy parameters for span-80 were previously developed, leading to simulated densities in agreement with experiment. The self-assembly behavior of span-80 in water was also different with the improved parameters. With the improved parameters, span-80 was found to form curved bicelles (bilayer with closed edges) in water (Figure B), in difference to the original parameterization that only predicts spherical micelles. This structure is a precursor to vesicles, which have been observed for span-80.4 The simulation was terminated after 200 ns, and it was determined that the time and length scales for the self-assembly of span-80 are beyond the capability of atomistic simulations. A periodic span-80 bilayer will therefore be used to study drug molecule interactions as a model for a vesicle patch. This project advances milestones ii and iii.

Figure A. Initial and final molecular dynamics simulation snapshots of linear amphiphilic delivery vehicles in water. Similar conformations are observed for cyclic amphiphilic delivery vehicles.

Hybrid MD-continuum simulation methodology for biomolecular systems: [D. Nikitopoulos (LSU), D. Moldovan (LSU), K. Hesary (LSU), B. Novak (LSU)]. Work has been conducted towards further development of formalisms and computational tools for coupling continuum CFD and MD simulation codes for study of non-equilibrium flow phenomena in mixed-scale domains involving biological materials. Kasra Hesary, working with the group, has successfully used these tools to couple LAMMPS to a commercial continuum code (ANSYS/ Fluent) and verified the results for simple test problems. The hybrid simulation implementation was tested using sudden-start Couette flow in single and two-phase systems, demonstrating the capability of these hybrid tools to yield quantitative predictions in agreement with analytical solutions. Currently we are working with CTCI colleagues to leverage use of CPU/GPU computing to further accelerate these simulations. This project advances milestone iii.

Polymeric nanoparticle-cell interactions [C. Sabliov (LSU-Ag), S. Grayson (Tulane), D. Moldovan (LSU), and R. Devireddy (LSU)]: The entrapment of bioactive compounds in polymeric nanoparticles enhances cellular uptake, drug efficacy, and drug stability. Particle size (1nm - 1mm), surface charge, and hydrophobicity play a vital role in the ability of nanoparticles to be endocytosed by the cell. Poly (lactic-co-glycolic) acid (PLGA) and chitosan-coated PLGA particles (PLGA/Chi) were developed to utilize the mucoadhesive property of chitosan and PLGA’s ability to efficiently entrap hydrophobic and hydrophilic drugs. PLGA and PLGA/Chi NPs loaded with antioxidants were synthesized and characterized. Particles measured 120 nm on average, with 100% entrapment efficiency. TEM clearly indicates PLGA particles aggregate under acidic conditions, while Chi/PLGA particles aggregate near neutral conditions. Loss of the positive charge of Chi/PLGA NPs near the pKa of chitosan (~6.5) adversely affects their mucoadhesive capacity and uptake. To ensure a positive charge of the nanoparticles under neutral physiological conditions, a chemically modified chitosan soluble at basic pH was synthesized and particles was made. The effect of size and charge on NP uptake will be assessed in Caco-2 cells over the next year. This project advances milestone iv.

Strategies for free energy evaluation in biomolecular systems: [D. Mobley (UNO), S. Rick (UNO), J. Baker (UNO), and H. Ashbaugh (Tulane)]. Over the past year an automated method for planning relative free energy calculations has been developed, enabling high throughput free energy estimates for as many as tens of thousands of systems. This work has been presented at national and international meetings, and this tool will be publicly distributed. We also have an ongoing collaboration within SD3 on improving sampling of biomolecular interactions, and are applying a new weighted ensemble technique from collaborators at the University of Pittsburgh to these problems. We continue to work on improving free energy techniques for molecular design in the context of biomolecular interactions, testing these techniques with applications to DNA gyrase and trypsin. This project advances milestone iii.Changes in research direction, future workAnne Robinson’s addition has initiated new research on tau-protein fibrogenesis.5 In collaboration with Ashbaugh, an REU student has been recruited in the summer of 2012 to explore this direction. This new project is an addition to SD3’s ongoing efforts in drug delivery vehicles.

Figure B. Simulation snapshots of a single span-80 monomer and a self-assembled span-80 bicelle.

References1) S. W. Rick “Replica exchange with dynamical scaling,” J. Chem. Phys., (2007), 126, 054102.2) B. A. Laurent and S. M. Grayson, “Synthesis of cyclic amphiphilic homopolymers and their potential application as polymeric micelles,” Polym. Chem. (2012), in press.3) Y. Li, M. D. Giles, S. Liu, B. A. Laurent, J. N. Hoskins, M. A. Cortez, S. G. Sreerama, B. C. Gibb, and S. M. Grayson “A versatile and modular approach to functionalisation of deep-cavity cavitands via “click” chemistry,” Chem. Commun., (2011), 9036-9038.4) K. Kato, P. Walde, N. Koine, S. Ichikawa, T. Ishikawa, R. Nagaharna, T. Ishihara, T. Tsuji, M. Shudou, Y. Omokawa, and T. Kuroiwa, “Temperature-sensitive nonionic vesicles prepared from Span 80 (sorbitan monooleate),” Langmuir, (2008), 24, 10762-10770.5) C. Ballatore, V. M.-Y. Lee, and J. Q. Trojanowski, “Tau-mediated neurodegeneration in Alzheimer’s disease and related disorders,” Nature Reviews, (2007), 8, 663-672.