Simulation Studies of Biomolecules @ Soft Interfaces :

1. Simulation Studies ofBiomolecules @ Soft Interfaces : Continuing challenge of bridging length- & time-scales Third Computational Chemistry Conference on “use of computational techniques in chemistry, biology, biochemistry, and materials science” SURA NCSA UoK ARL www.2003

2. SURA-NCSA-UoK-ARL CCC2003* It gives me great pleasure to talk to you today about research going on in my group at UPenn that is focused on membranes and membrane-bound species. * DISCLAIMER - This talk contains no equations – theoreticians may find the extensive use of color images offensive

3. PROLOGUE In the 1960’s COMPUTATION started to become respectable…. My early computations dealt with ATOMIC systems. But, in the mid-1970’s, computers became more powerful I shifted to MOLECULAR solids & liquids. Since the 1980’s my research has dealt with simulation algorithms and applications ranging from materials to bio-membranes. Current interests - ab initio based (DFT) methods forchemical & enzyme reactions and coarse-grain simulations of self-assembling soft matter systems.

4. Use computer to follow the motion of a system of atoms or molecules using principles of physics (Newton, Lagrange, Feynman) and statistical mechanics (Onsager, Kubo, van Hove) to go from trajectories to observables. Interactions can be from empirical potentials or from quantum mechanics, often via DFT - Car-Parrinello methodology. Methodology

5. Biomolecules @ Soft Interfaces* Carlos Lopez, Steve Nielsen, Preston Moore, Robert Doerksen, Srinivas Goundla, Rosalind Allen, Bin Chen, John & Mee Shelley * Bridging from atoms to the mesoscale

6. Infection Pathway of a Virus in a Living Cell* Single-molecule real-time visualization of the infection pathway of single viruses in living cells, each labeled with only one fluorescent dye molecule. Trajectories show various modes of motion of adeno-associated viruses (AAV) during their infection pathway into living HeLa cells: • Consecutive virus touching at the cell surface and fast endocytosis; • Free and anomalous diffusion of the endosome and the virus in the cytoplasm and the nucleus; and • Directed motion by motor proteins in the cytoplasm and in nuclear tubular structures. * Munich Group Science

7. Polymersomes (Penn MRSEC)* *Discher, Hammer, Bates (Science 1997)

8. Nature’s Nanoworld * *Cell membranes are complex, containing lipids, proteins, cholesterol, carbohydrates, plus actin filements, etc… Design principles for Nature’s devices, plus their self- & supramolecular assembly can yield new materials.

9. Membranes – A Playground for Molecular Dynamics Simulation • MANY ACHIEVEMENTS • MANY ACTIVE GROUPS: • Schulten, Carloni, Sansom, • Tielman, Roux, Pastor, etc.

10. Ion Channel in a Lipid Bilayer – n-AChR (M2-TM) • M2 pentamer in a lipid bilayer at T=303K • 94 DMPC lipids plus5 M2-TM peptides Channels well-studied • Schulten • Roux • Carloni • Sansom • Teilman N-terminal intracellular on top

11. Nature’s Ion Channels • Transmembrane – Assembly & Function Difficult with classical MD • Mostly pre-assembled systems • Simulated structures Open/Closed? • Gramicidin (Warshel, Roux – Karplus,…) • K+ (Carloni, Roux, Tielman, Sansom,…) • OMP, Porins (Carloni, Sansom, Schulten,…) Design Principles ?

12. Carbon Nanotube as Nanosyringe ? G. Hummer, J. Rasaiah, & J.P. Noworyta Nature 414, 188-190 (2001) Water in a carbon nanotube Tube:13.4 Å long 8.1 Å diameter “We observe pulse like transmission of water through the nanotube…two-state transitions between empty and filled states on nanosecond timescale..”

13. Why Coarse Grain ? * Atomistic Models • Empirical potentials - local structures • Limited system size • Limited timescale *One million lipids / m2

14. A Coarse-grain Model for Soft Interfaces Strategy for Simulations of Advanced Materials

15. A Coarse-grain Modelfor Soft Interfaces* Carlos Lopez, Steve Nielsen, Preston Moore, Robert Doerksen, Srinivas Goundla, Rosalind Allen, Bin Chen, John & Mee Shelley * Bridging from atoms to the mesoscale

16. From hundreds of atoms to billions - the challenge of the mesoscale Coarse grain models & soft interfaces

17. Coarse Grain DMPC Choline Head Group Phosphate Group Glycerol Groups Acyl Chains Chain Ends

18. Self Assembly of DMPC-CG • NVT Ensemble:46 Å X 45 Å X 59 Å • 64 DMPC,T: 30°C, 1ns (20fs/step)

19. Coarse Grain Membrane - Results J Phys Chem 2001 MUCH faster than atomistic models See JPC 2001

20. DMPC-CG 1024 Bilayer • NPT Ensemble • Orthorhombic Cell • 1024 Lipids • 1 ns (20fs/step) • T: 30°C, P: 1atm • Equilibration  340ps Snapshots of Configurations Comp Phys Comm 2002

21. Membrane Surface Dynamics • Area: 18nm X 20nm • Note Transient Holes • P(Blue), N(Orange) Water Removed for Clarity

22. Challenges & Opportunities Samuel Johnson: 1709-1784 • Nothing amuses more harmlessly than computation, and nothing is oftener applicable to real business and speculative inquiry. • A thousand stories which the ignorant tell and believe die away at once when the computist takes them in his grip.

23. Nanosyringe - Nature’s Design Protein Helix Bundle Coarse Grain Tube Hydrophobic Pore diameter sized to accommodate “water”

24. Nanosyringe • CG-tube mimic of nanotube or viral channel insertion into membrane • NPT-MD 30°C, 1atm

25. Transmembrane Nanotube • CG-tube mimic of membrane-bound nanotube or viral channel • NPT-MD 30°C, 1atm

26. Hydrophobic Nanotube CG nanotube (nanosyringe) in CG DMPC • Tube gets blocked with lipids after 15,000 steps • First lipid inserted at 5000 steps (10 ns ?) • Tube tilts to fit bilayer • Q: How to make a nanosyringe…? • A: Nature uses hydrophilic caps

27. Nanosyringe - Nature’s Design Protein Helix Bundle Coarse Grain Tube Hydro-phobic Hydro - philic Pore diameter sized to accommodate “water”

28. Insertion into Membrane Tube inserts into the membrane, favoring hydrophobic interactions Start outside the membrane (water not shown)

29. Insertion into Membrane Tube drags lipids into the middle of the bilayer. Then “sees” the other side of the membrane. Tube straightens up and then remains in this position through the run.

30. Nature’s Design CG tube in DMPC: • “Water” goes through tube • NPT 30°C, 1atm • Poreation Snapshot of tube in CG DMPC Two waters are present in the tube. Lipids removed from bottom leaflet to reveal the tube/lipid interface

31. Insertion into Membrane

32. Antibacterial Peptide Molecules • Common traits of AB peptides: • Relatively short peptides. • Charged and hydrophobic groups segregate onto opposite sides of a structure. • Believed to kill cells by disrupting membranes. Q: De novo design of biomimetic antimicrobial molecules?

33. Anti-microbial Peptide Mimics Bin Chen, Carlos Lopez, Robert Doerksen, Bill DeGrado

34. Magainin Mimic Polyarylamide n = 3 1

35. de novo design and synthesis of antimicrobials • Gradient-corrected density functional theory calculations • Atomistic molecular dynamics and Monte Carlo calculations • Coarse-grained molecular dynamics simulations

36. A schematic flow-chart showing the whole design process Search low-energy conformations, parameterize torsion potentials Build initial target polymer backbones Density functional theory calculations Monte Carlo and molecular dynamics Test force fields, investigate the polymer’s conformations at vacuum and interfaces and its hydrophobicity, select the potential target Experimental synthesis and assay tests molecular dynamics simulations at membrane/water interface Evaluate the polymer’s antibacterial activity and other important properties, determine the final target Provide microscopic-level insights on the antibacterial mechanisms and guidelines for improving the antibacterial polymer

37. DeNovo Designed Anti-microbial Polymers PNAS 2002

38. Antibacterial Activity of Polyamides n MIC E. coliK. pneumoniaeB. subtilis 2 19 66 12 3 <19 N/A 19 4 7.5-15 31-50 16 6 >500 250 >500 (AB)n

39. A Coarse-grain Simulation Model for Probing Mechanisms of Anti-bacterial Action* Srinivas Goundla, Carlos Lopez, Steve Nielsen Michael L. Klein* Center for Molecular Modeling University of Pennsylvania

40. CG-AB with CG Lipid Approximation to AB: • Use existing CG types to emulate the AB molecule.

41. AB Dimer in CG Lipid Peptide mimics are adsorbed at the lipid surface. Peptide mimics first enter bilayer and eventually settle under the head groups

43. Acknowledgements • Thanks to friends & collaborators • NPACI @ NCSA, SDSC, PSC • NSFNIH

44. Many thanks for inviting me to talk to you!

45. END

46. Challenges & Opportunities Terascale Computing has arrived • It will surely come to the desktop… • Phenomena at longer-length and time scales will be accessible… • HPC will participate in discovery of advanced materials and more…

47. Challenges & Opportunities Samuel Johnson: 1709-1784 • Nothing amuses more harmlessly than computation, and nothing is oftener applicable to real business and speculative inquiry. • A thousand stories which the ignorant tell and believe die away at once when the computist takes them in his grip.