1997 MRS Fall Meeting & Exhibit

Chairs

Bela Mulder, FOM Inst
Christoph Schmidt, Univ of Michigan
Viola Vogel, Univ of Washington 

Symposium Support 

• The Whitaker Foundation

Proceedings published as Volume 489 
of the Materials Research Society 
Symposium Proceedings Series.

* Invited paper

SESSION K1: MOTILITY AND FORCE GENERATION 
Chairs: Frederick C. MacKintosh and Christoph F. Schmidt 
Monday Morning, December 1, 1997 
Grand Ballroon (S)

8:45 AM *K1.1 
USING OPTICAL TWEEZERS TO STUDY KINESIN MOTORS. Steven Block, Princeton University, Princeton, NJ.

Abstract not available

9:15 AM *K1.2 
MECHANICAL PROPERTIES OF CYTOSKELETAL PROTEIN. J. Howard, University of Washington.

Abstract not available

9:45 AM K1.3 
STATISTICAL THEORY OF MUSCLE CONTRACTION: NEW ANALYTIC RESULTS VERSUS EXPERIMENTS. Imre Derényi, Tamás Vicsek, Dept. of Atomic Physics, Eotvos University, Budapest, HUNGARY.

Myosin is an ATPase enzyme that converts the chemical energy stored in ATP molecules into mechanical work. During muscle contraction, the myosin cross-bridges attach to the actin filaments and exert force on them yielding a relative sliding of the actin and myosin filaments. In this paper we present a simple mechanochemical model for the cross-bridge interaction involving the relevant kinetic data and providing simple analytic solutions for the mechanical properties of muscle contraction, such as the force-velocity relationship or the relative number of the attached cross-bridges. So far the only analytic formula which could be fitted to the measured force-velocity curves has been the well known Hill equation containing parameters lacking clear microscopic origin. The main advantages of our new approach are that it (i) explicitly connects the mechanical data with the kinetic data and the concentration of the ATP and ATP-ase products, (ii) and as such it leads to new analytic solutions which agree extremely well with a wide range of experimental curves, while the parameters in the corresponding expressions have well defined microscopic meaning.

10:30 AM *K1.4 
TORQUE GENERATION BY THE FLAGELLAR ROTARY MOTOR. Howard C. Berg, Departments of Molecular & Cellular Biology and of Physics, Harvard University, Cambridge, MA.

The rotary motor that drives a flagellar filament of the bacterium Escherichia coli is only about 40 nm in diameter, yet contains about 20 different kinds of parts. It spins alternately clockwise or counterclockwise at speeds in excess of 100 Hz. It is powered by a proton electrochemical gradient. A number of its dynamic properties will be discussed.

11:00 AM *K1.5 
FORCE PRODUCTION BY ASSEMBLING MICROTUBULES. M. Dogterom, F.O.M. Institute AMOLF, Amsterdam, NETHERLANDS.

The assembly and disassembly of cytoskeletal filaments can generate forces that are important for various forms of intra-cellular motility. For instance, forces generated by microtubules (MTs) are believed to contribute to the correct positioning of chromosomes during the various stages of cell division. In fact, MTs can exert both pulling and pushing forces on chromosomes while switching between growing and shrinking states (a behavior termed `dynamic instability'). I will present experimental results on the force that can be produced by the growth of a single MT under simplified in vitro conditions, and show how an opposing force effects the assembly rate of the MT (the `force-velocity curve').

11:30 AM K1.6 
MECHANOCHEMICAL SPRINGS. L. Mahadevan, M.I.T.; P. Matsudaira, Whitehead Institute and M.I.T.

Force generation at the sub-cellular level occurs generally in one of two ways: by the continuous transduction of chemical energy (ATP hydrolysis) or the chemically mediated release of stored mechanical energy, as in the release of a spring. The latter is common in many invertebrates such as the acrosomal reaction of the Limulus sperm or in reversible stretching of the stalk of peritrich ciliates such as the Vorticella, first studied by Leeuwenhoek more than two centuries ago. In both these systems, motility is driven reversibly by divalent ions such as Ca++. We study the dynamics of the acrosomal process, where the structure of the acrosome, an actin bundle, and the conformation changes in it during the uncoiling process are known. Following various independent estimates of the energetics of the process, we construct a model for the acrosomal process in terms of the variation of the geometry of the bundle. This is defined by an order parameter that takes into account its curvature and torsion. Then the dynamical evolution of the shape is determined by the propagation of a mechanical phase boundary along the bundle that is driven by a change in the chemical potential due to the presence of Ca++ ions. Calorimetric measurements give us an idea of the various energy levels in the energy landscape. Based on this, some experimental predictions for the acrosomal process are given. Finally, some implications of the model for the general problem of the dynamics of protein polymorphism, a subject of much interest, are outlined.

11:45 AM K1.7 
QUANTAL SHORTENING IN MUSCLE SARCOMERS. Gerald H. Pollack, Dept. of Bioengineering, Univ. of Washington, Seattle, WA.

Viewed from a materials perspective, muscle is a gel-like organelle built of a parallel array of cross-linked polymer strands. Its function is to convert chemical into mechanical energy. The process by which this occurs is unsettled and the field is ripe for fresh, engineering-oriented approaches. We focus on the single sarcomere, the smallest (micron scale) functional unit that retains the natural polymeric structure. Tension is measured using nanofabricated cantilever transducers and length dynamics are measured from high-resolution optical images of the single sarcomere. We find that the sarcomere shortens (and relengthens) in discrete steps. Periods of pause alternate with steps of rapid shortening. Step-size is not constant; rather, the distribution is broad. However, the distribution function comprises multiple uniformly spaced peaks, implying that the size of each step is a multiple of a quantum value. The best-fit quantum is approximately 2.5 nm. Typical step-size values fall in the range 10 - 15 nm (i.e., 4, 5, or 6 x 2.5 nm). Thus, the array of polymers collectively shortens (and lengthens) by integer multiples of 2.5 nm. This quantum value has several structural counterparts that may be relevant in producing the step. One source may be the unfolding of a single turn (2.5 nm) of the beta-sheet, the latter appearing in multiple copies in a string-like array along one of the polymers (titin). Another source may be the sliding of alpha-helical coiled-coils with 2.5 nm linear surface-charge repeats, past one another. A third is the sliding of one polymer strand (actin) past another (myosin), the former having binding sites for the latter distributed every 2.7 nm. Any/all of these mechanisms could underlie both the quantized steps and ultimately the mechanism of contraction.

SESSION K2: CYTOSKELETAL POLYMERS 
Chairs: Marileen Dogterom and Paul A. Janmey 
Monday Afternoon, December 1, 1997 
Grand Ballroon (S)

1:30 AM *K2.1 
OPTICAL PROBES OF RHEOLOGY: FROM GELS TO CELLS. T. Gisler, P. Kaplan and D.A. Weitz, Dept. of Physics, University of Pennsylvania, Philadelphia, PA.

The thermal motion of small particles can often reflect the properties of the local environment of the particle. The dynamics of these probe particles can be measured using light scattering, or directly using imaging. This provides a potential of measuring local rheological properties using non invasive optical means. Here we discuss developments of this method through the use of light scattering to probe the local rheology. We review a simple interpretation of dynamic light scattering data that provides a measure of the frequency dependent viscoelastic moduli of the medium in which the probe particle is embedded. The applicability of this method is tested using a both gels, and entangled networks, of polymers and biopolymers. We also discuss the application of this method to directly probe the local rheology within cells by monitoring the motion of beads inside the cells using a microscope configured to both image the particles, and to detect the light scattered from them.

2:00 PM *K2.2 
EXPOSING MECHANICAL PROPERTIES OF HIDDEN STRUCTURAL COMPONETNS INSIDE CELLS. Evan Evans1,2 and Micah Dembo2, 1Physics, Univ. of British Columbia, Vancouver, CANADA; 2Biomedical Eng., Boston Univ, Boston, MA.

Biophysicists, bioengineers, and biologists have been pushing and pulling on cells for nearly a century in order to deduce mechanical properties that emanate from subcellular structures. Frustrating this goal is the heterogeneity in cell structure where several solid-like components are embedded in a fluid-like cytosol and may or may not be linked ultimately to the fluid bilayer boundary of the cell. As such, each component moves differently under large changes in cell shape. Because of this ''incommensurability'', we need to expose the separate deformation response of each element in mechanical tests of cells. Called FIMD (luorescnece maging under icromechanical eformation), we developed a method that combines differential fluorescence labeling and micromechanical instrumentation to ''map'' local density fields of distinct components under controlled cell extensions. Originally used to examine the membrane cytoskeleton of the red blood cell(Discher, Narla, Evans, Science 266, 1994), we demonstrate here that sequences of density maps of structures in more complex cells can be analyzed to obtain the precise deformation state of local elements of an interior component as cell deformation progresses. These deformation fields can be compared with computational simulations of the structure subjected to the same macroscopic shape changes for different physical models. The values of material coefficients in successful models are then found from the levels of mechanical force used to deform the cell.

2:30 PM K2.3 
MICROTUBULE AND ACTIN FILAMENT DYNAMICS IN LIVING CELLS: OBSERVATIONS WITH THE NEW POL-SCOPE. R. Oldenbourg, K. Katoh, P. T. Tran, Marine Biological Laboratory, Woods Hole, MA; C. C. Hoyt, Cambridge Research and Instrumentation Inc., Cambridge, MA.

The living cell is criss-crossed by dense networks of filaments providing mechanical stability, site directed molecular and organellar transport and support of other vital cell functions. Filaments in the living cell are constantly assembled and disassembled and the network architecture changes depending on the state of the cell. Because of the pervasive nature of filaments in the living cell, their reorganization into different networks, and their vital role in cell functions, the visualization of the dynamic network architecture is very important for understanding the molecular biology of the cell. With the polarized light microscope we can observe the birefringence associated with thin filaments or partially oriented filament networks and measure the birefringence directly in living cells. Filament birefringence is a consequence of the elongated shape of the molecules and occurs naturally without the need to stain or label them, as is necessary in fluorescence imaging. We have measured the birefringence inside living cells using the new polarized light microscope (Pol-Scope; 1, 2). The design of the Pol-Scope is based on the traditional polarized light microscope in which the crystal compensator is replaced by a precision universal compensator made from two liquid crystal variable retarders. Electronic image acquisition in the Pol-Scope is synchronized to liquid crystal settings capturing a sequence of four images with circular and elliptical polarization. The recorded images are used to calculate specimen retardances, measuring the magnitude of retardance and its slow axis direction in each image point. Complete retardance images are recorded at 3 second time intervals compiling a dynamic data record at high spatial and temporal resolution. These capabilities significantly enhance the analytic power of the polarizing microscope in all its traditional application areas, ranging from measuring strain in industrial material samples to minute birefringences of individual biopolymers. We will present measurements of single filament retardances and show time lapse movies of retardance images documenting the dynamic network architecture of systems such as microtubules in dividing epithelial cells, and actin filaments in growth cones of living nerve cells.

2:45 PM K2.4 
INTERACTIONS BETWEEN CYTOSKELETAL FILAMENTS. Jagesh V. Shah+, Lisa A. Flanagan, Josef Kaes* and Paul A. Janmey, Division of Experimental Medicine, Brigham and Women's Hospital, Boston, MA; + Harvard-MIT Division of Health Sciences and Technology, Boston, MA; * Department of Physics, University of Texas, Austin, TX.

The cytoskeleton is an intracellular network of protein filaments which, amongst other functions, provides mechanical integrity to the cell. The cytoskeleton is comprised of three chemically distinct polymers: microfilaments (F-actin), microtubules and intermediate filaments. Bulk rheologic and single filament measurements of each filamentous system indicate that each type of cytoskeletal polymer contributes distinct mechanical properties to the cell. Experimental manipulation of one cytoskeletal protein affects the organization of the others, suggesting that the three cytoskeletal filaments types interact in vivo, and recent findings by several groups have identified putative linking proteins between specific pairs of filaments. The mechanical consequences of an intersystem crosslink depends on the geometry and the stability of the crosslink formed. We have used microscopic methods to visualize the reptation of individual cytoskeletal filaments in a heteropolymeric cytoskeletal gel in the presence of putative intersystem linkers. Using this method we hope to detect changes in reptation dynamics that would indicate a direct interaction between two different cytoskeletal filaments mediated by a crosslinking protein. Preliminary results indicate that MAP2c, the microtubule associated protein and microfilament gelation factor found in neurites and neuronal growth cones, crosslinks microtubules to microfilaments. These results indicate that MAP2c may play a novel role in the cytoskeletal organization and mechanical properties of the neuronal growth cone where microtubules, microfilaments and MAP2c co-localize. Furthermore, visualization of reptation in heteropolymeric gels may provide insight into the dynamics of composite polymeric materials and their dependence on filament stiffness and matrix rheologic properties.

3:30 PM *K2.5 
VISCOELASTICITY AND ITS MICROSCOPIC CHARACTERIZATION IN SEMIFLEXIBLE BIOPOLYMER SOLUTIONS. F.C. MacKintosh, F. Gittes, B. Schnurr, P.D. Olmsted*, and C.F. Schmidt, Univ of Michigan, Dept of Physics and Biophysics Research Div, Ann Arbor, MI; *Univ of Leeds, Dept of Physics, Leeds, UNITED KINGDOM.

Plant and animal cells contain a complex polymeric network known as the cytoskeleton. A principal component of this is the actin cortex, a gel-like network of F-actin protein filaments. Recently, solutions of reconstituted F-actin have provided in vitro models of the actin cortex, as well as excellent model systems in which to study semiflexible polymers. We describe models of viscoelasticity in semiflexible polymers, and report theoretical and experimental results for thermal fluctuations of embedded particles, which act as local viscoelastic probes of soft materials such as biopolymer solutions. Specifically, we report scaling of the shear modulus with frequency, in contrast with the behavior of flexible polymer systems.

4:00 PM K2.6 
STATISTICAL MECHANICS OF SEMIFLEXIBLE POLYMERS AND MACROMOLECULAR NETWORKS. Erwin Frey, Technische Universitaet Muenchen, Garching, GERMANY.

We study the physical properties of macromolecular networks (e.g. cytoskeleton), consisting of semiflexible polymers such as actin. We start by giving a theoretical analysis of the conformational statistics and mechanical response of single filaments. Experimentally relevant quantities such as the end-to-end distribution function and the force-extension relation are discussed. The results are used to analyze the elastic modulus of an entangled solution of such objects. We also discuss the short time and long time (terminal regime) dynamics and compare our results with recent dynamic light scattering and viscoelastic measurements. Furthermore, we analyze the elasticity transition in stochastic models of crosslinked networks.

4:15 PM K2.7 
MODEL FOR DYNAMIC SHEAR MODULUS OF SEMIFLEXIBLE POLYMER SOLUTIONS. F. Gittes, B. Schnurr, C.F. Schmidt, P.D. Olmsted*, and F.C. MacKintosh, Univ of Michigan, Dept of Physics and Biophysics Research Div, Ann Arbor, MI; *Univ of Leeds, Dept of Physics, Leeds, UNITED KINGDOM.

We discuss a dynamical model for the frequency-dependent shear modulus of an entangled solution of semiflexible polymers, based on longitudinal fluctuations in filaments between entanglement points or crosslinks. The goal is to explain non-Rouse, power-law scaling of the bulk shear modulus that is found via microscopic rheology of highly entangled F-actin solutions. This generalizes a previous model for the static modulus. Hydrodynamic effects, and the validity of a local drag approximation below the scale of the mesh size, are discussed. We test aspects of our model via a molecular dynamics simulation, and also present for comparison experimental results from microrheology on F-actin.

4:30 PM K2.8 
ELECTROSTATICALLY INDUCED BUNDLE FORMATION OF RODLIKE POLYELECTROLYTES. COMPARISON OF PREDICTIONS FROM MONTE CARLO SIMULATIONS WITH EXPERIMENTS ON FD AND M13 VIRUS PARTICLES. Lars Nordenskiold, Alexander Lyubartsev, Stockholm University, Arrhenius Laboratory, Stockholm, SWEDEN; Jay Tang and Paul Janmey, Div. of Experimental Medicine, Brigham and Women's Hospital; Boston, MA.

Many filamentous biopolymers such as F-actin, microtubules, tobacco mosaic virus and the bacteriophage fd virus form bundles under appropriate solution conditions. All these macromolecules are cylindrical (rodlike) negatively charged polyelectrolytes. The association of the macromolecules into bundles is induced by different multivalent cations and the behaviour is analogous to the condensation of DNA under similar circumstances. The present work concerns the bundling of the fd virus particle and its mutant M13 which has a reduction of axial charge density by a factor of 3/4 as compared to the wild type fd. The electrostatic force (or osmotic pressure) between these ordered polyelectrolytes, as a function of interaxial separation, has been calculated with Monte Carlo simulations, using a method previously applied to ordered DNA (Lyubartsev and Nordenskiold, J. Phys. Chem., 1995, 99, 10373). The theoretical simulations have been compared with experiments on the bundling of fd and M13 caused by different multivalent ions, as detected by light scattering and electron microscopy. In the theoretical calculations, the bundling is caused by electrostatic attraction between the neighbouring polyelectrolytes, caused by the correlated interactions between the ion clouds belonging to different charged macromolecules. The importance of this effect (i. e. capacity to induce bundling) is governed by axial charge density of the polyelectrolyte, amount of multivalent ion present, charge and size of the hydrated multivalent ion. These predictions are in qualitative agreement with the experimental results on bundling of fd and M13. It is suggested that electrostatic attraction is a general mechanism for stabilising bundles of filamentous biopolyelectrolyte systems.

4:45 PM K2.9 
TACTOIDAL GRANULES IN CONCENTRATED ACTIN GELS-A COLUMNAR STATE OF PROTEIN FILAMENTS. Jay X. Tang, Philip G. Allen, Paul A. Janmey, Division of Experimental Medicine, Brigham & Women's Hospital, Boston, MA; Rudolf Oldenbourg, Marine Biological Laboratory, Woods Hole, MA.

Actin is an abundant cytoskeletal protein in most mammalian cells. The polymerized filament, F-actin, is a semiflexible biopolymer of 8-10 nm in diameter and has a persistence length of a few microns. We have observed in vitro spindles of F-actin, imbedded in a background suspension of orientationally ordered actin filaments, i.e., a nematic state of F-actin. Phase contrast microscopy shows these stable spindles of densely packed F-actin to be of various size in the order of 10 m. The aspect ratio of long to short axes of the spindles depends on the average length of F-actin, and is approximately 5-10 for a mean filament length of 2 m. These granules demonstrate elastic response upon micro-manipulation. The well defined interface and smooth curvature suggest adjustments of filament positions in the formation of these spindles. Measurements of local birefringence by a pol-scope reveal that the optical birefringence is largely suppressed within the granules compared to the surrounding nematic background. Parallel alignment of filaments along the long axes of tactoids are detected within the granules by electron microscopy. We propose that these spindles of F-actin are domains of columnar liquid crystalline phase, co-existing with the nematic state of F-actin of a few percent in volume fraction. Further characterization of interfilament interactions of F-actin in this physical state may help understand the structural roles of actin in cytoskeleton.

SESSION K3/L3: JOINT SESSION: 
PROTEINS: ADSORPTION AND INTERACTIONS WITH MEMBRANES 
Chairs: Robert S. Cantor and Igal Szleifer 
Tuesday Morning, December 2, 1997 
Grand Ballroon (S)

8:30 AM *K3.1/L3.1 
PROTEIN ADSORPTION ON SURFACES WITH GRAFTED POLYMERS. I. Szleifer, Department of Chemistry, Purdue University, West Lafayette, IN.

The prevention of protein adsorption on surfaces in one of the important requirements in the design of biocompatible materials. In recent years it has been shown that grafting of polymer molecules to the surface of the biomaterial may be very effective in preventing protein adsorption. In this talk the factors that determine the ability of the grafted polymers to prevent protein adsorption will be discussed. A systematic theoretical and experimental study of the ability of grafted poly-ethylene oxide (PEO) to prevent the adsorption of lysozyme and fibrinogen will be shown. The theoretical studies are based on the application of single-chain mean-field theory. The predictions of the theory are in very good agreement with the experimental observations. Furthermore, the theory provides a microscopic understanding of the mechanism of protein adsorption (and/or rejection) on surfaces with grafted polymers. It is found that the most important parameter in controlling protein adsorption is the polymer surface coverage, the polymer chain length has only a minor effect. It is found that in hydrophobic surfaces the grafted PEO also adsorbs to the surface and thus, it prevents protein adsorption by covering adsorbing sites for the proteins. In the cases of surfaces that PEO does not adsorb, like PEO-lipid liposomes, the polymer is less effective in preventing protein adsorption and the mechanism is based on steric repulsion. The time scale for the proteins to reach the surface is predicted to be much faster in hydrophobic surfaces than for the surfaces that do not attract the PEO segments. The effect of changing the structure of the polymer molecules in preventing protein adsorption will be shown for a variety of polymer chemical architectures.

9:00 AM K3.2/L3.2 
ENHANCED BLOOD COMPATIBILITY OF SILICON SURFACES COATED WITH SELF-ASSEMBLED POLYETHYLENE GLYCOL FILMS. Miqin Zhang, Mauro Ferrari, University of California at Berkeley, Biomedical Microdevices Center, Department of Materials Science and Mineral Engineering, Berkeley, CA.

The covalent attachment of Self-Assembled (SA) polyethylene glycol (PEG) coating on silicon surfaces was investigated. The PEG coatings were immobilized by the introduction of silanol groups on the PEG chain ends, which react with the hydroxyl groups on silicon surfaces. The coated films were characterized by contact angle measurement, ellipsometry, and electron spectroscopy for chemical analysis (ESCA). The interaction between treated surfaces and blood components was investigated to assess the performance of the modified surface. The protein adsorption and platelet adhesion were investigated by ellipsometry and optical microscopy, respectively. The PEG modified silicon surfaces showed excellent blood compatibility, including both the reduction of plasma protein adsorption and the suppression of platelet adhesion.

9:15 AM K3.3/L3.3 
PLASMA PROTEIN ADSORPTION AND PLATELET ADHESION ON POLY[BIS(TRIFLUOROETHOXY)PHOSPHAZENE]. Michael Grunze, Alexander Welle, University Heidelberg, Institute of Applied Physical Chemistry, Heidelberg, GERMANY; Dsidra Tur, Academy of Sciences, Institute of Organoelement Compounds, Moscow, RUSSIA.

Poly[bis(trifluoroethoxy)phosphazene] (PTFEP) is a biocompatible material [1] used as bulk material in medical implants. We developed a process to coat surfaces with PTFEP films and performed ELISA experiments designed to understand their blood compatibility. Coatings with different chemical properties (hydroxylated glass, amino-, methyl- and aldehyde- terminated silane films) were used as reference materials. We observed that PTFEP adsorbs preferentially Albumin from blood plasma and only small amounts of coagulation / inflammation stimulating proteins like Fibrinogen, von Willebrand Factor, Fibronectin and Immunoglobulines. In general, there is a good correlation between increasing coverage of adsorbed Albumin and reduced in vitro platelet adhesion. However, there is no direct correlation between the amount of Fibrinogen and platelet adhesion, restricting the use of Fibrinogen as a marker protein for platelet adhesion. An important prerequisite of blood compatibility is the stabilization of the native state of adsorbed proteins, since denaturated proteins stimulate platelet adhesion. Elution of adsorbed proteins by sodiumdodecylsulfate solution was used to quantify the amount of irreversibly attached and presumably denaturated proteins. PTFEP showed, compared to the reference materials, a low amount of irreversibly adsorbed proteins of the coagulation cascade. Circular Dichroism measurements of adsorbed Fibrinogen showed the same trend of protein denaturation at the surface. It was found that the secondary structure of adsorbed Fibrinogen is altered to a lower extent on PTFEP as compared to the reference materials [2]. We conclude that PTFEP has a unique blood compatibility because of the favorable composition and the stabilization of the adsorbed protein layer. We suggest that the trifluoroethanol moieties of PTFEP are responsible for these effects.

9:30 AM K3.4/L3.4 
MELTING AND INTERACTIONS IN MULTILAYERS OF TWO-DIMENSIONAL CRYSTALS OF MEMBRANE-PROTEIN BACTERIORHODOPSIN. I. Koltover, T. Salditt, J.O. Rädler, C.R. Safinya, Materials Department, Physics Department and Biochemistry and Molecular Biology Program, University of California at Santa Barbara, Santa Barbara, CA; K.J. Rothschild, Physics and Molecular Biophysics Laboratory, Boston University, Boston, MA.

Bacteriorhodopsin (bR), a light-driven proton pump, is an integral membrane protein of bacterium Halobacterium Salinarium. In the native bacterial membrane, bR self-assembles into regular hexagonal crystalline arrays in the plane of membrane. This two-dimensional (2D) protein crystal undergoes a fully reversible melting transition as a function of temperature. We have conducted a synchrotron x-ray diffraction study of oriented multilayers of bR-containing native bacterial membrane patches, as well as stacks of novel giant (50 diameter) single-crystal fused bR membranes. The precise in-situ control of humidity and sample temperature combined with the line-shape analyses of x-ray diffraction peaks allowed us to elucidate and control the intra- and inter-membrane protein-protein interactions. The important findings are as follows. First, the ordered 2D self-assembled lattice of proteins is found to exhibit diffraction patterns characteristic of a 2D solid with power-law decay of in-plane positional correlations, which allows to measure the elastic constants of protein crystal. Second, the melting temperature is a function of the multilayer hydration, with two distinct regimes of melting: lower hydration regime of orientationally commensurate membranes and high hydration regime of completely decoupled membranes. Third, preparation of nearly perfect (mosaicity<0.04circ) multilayers of fused bR membranes permitted, for the first time, application of powerful interface x-ray scattering techniques to a membrane-protein system, revealing the membrane corrugations and elucidating correlations of proteins at different hydration levels. Supported by NSF grant DMR 962091, and the Petroleum Research Fund (No.31352-AC7) to CS and NSF grant MCB-9419059 and ARO-SBIR/AmberGen contract to KJR.

9:45 AM K3.5/L3.5 
FLUCTUATIONS OF ACTIVE MEMBRANES. Jean-Baptiste Manneville, Patricia Bassereau, Jacques Prost, Physico-Chimie Curie UMR168, Institut Curie, Paris, FRANCE.

Incorporating pumps or ion channels into lipid bilayers allows us to study the out-of-equilibrium dynamics of fluctuating active membranes. Theoretical calculations (1,2) predict that fluctuations of such `active' membranes should differ both qualitatively and quantitatively from that of `passive' membranes, ie. membranes without any pumping activity. We observe the dynamical properties of fluctuating giant phospholipid visicles containing bacteriorhodopsin (BR), a light activated proton pump. Using micropipets techniques described by Evans et al. (3), we study the effects of BR incorporation and activation on the bending stiffness of the membrane. Modifications of the fluctuation spectrum upon BR activation is investigated by reflexion interference contrast microscopy (RICM) (4).

10:00 AM K3.6/L3.6 
OBSERVATION OF ION CHANNELS IN VESICLES UNDER CONTROLLED TENSION USING A MICROPIPET-ELECTRODE WITHOUT GIGA-SEAL. M. Gouliana, O.N. Mesquitab, D. Kuchnir Fygensona, E. Mosesb, C. Nielsenc, O.S. Andersenc and A. Libchabera,b. aRockefeller University, bNEC Research Inst., Princeton, NJ and cCornell University Medical College, Ithaca, NY.

We observe the effect of membrane tension on cross-membrane dimerization of the membrane-bound peptide gramicidin A. Gramicidin dimers form a simple ion channel when their chains ( 15 amino-acids) align tail-to-tail in the two leaflets of a bilayer membrane. The resulting pore is selective for monovalent cations. Kinetics of gramicidin pore formation have been studied extensively in different membranes, revealing a dependence on bilayer thickness. We observe gramicidin kinetics in large, single-component (DOPC), unilammelar vesicles using a micropipet without the standard ``giga-seal''. Absence of a giga-seal makes the membrane tension a well defined control parameter. We find the rate of channel formation increases with increasing tension in the membrane. We interpret this as a consequence of tension induced thinning of the bilayer, since the length of a gramicidin dimer is less than the relaxed bilayer thickness.

10:45 AM K3.7/L3.7 
LATERAL PRESSURES IN CELL MEMBRANES: MODULATION OF PROTEIN FUNCTION BY VARIATION IN LIPID COMPOSITION. Robert S. Cantor, Dartmouth College, Dept of Chemistry, Hanover, NH.

Variations in the composition of cell membranes can strongly influence the function of proteins embedded therein. However, in most cases it is not known whether lipids and other membrane components act by binding directly to proteins, or indirectly through changes in a structural or thermodynamic property of the fluid bilayer. In the present work, a simple thermodynamic analysis is developed based on the hypothesis that variations in membrane composition induce changes in the distribution of transverse pressure in lipid bilayers. If protein function involves a conformational transition accompanied by a depth-dependent change in its cross-sectional area, it is predicted that a small redistribution of lateral pressures can induce a large shift in the protein conformational equilibrium. The sensitivity of this equilibrium to the lateral pressure profile arises predominantly from the localization of the large interfacial free energy within a domain of molecular thickness. Lattice statistical thermodynamic calculations are used to predict the effect of cholesterol and lipids of different length, stiffness and head groups on the pressure profile. Of particular interest are small interfacially-active solutes, as a model for general anesthetics: a strong correlation is found between anesthetic potency and increased lateral pressure near the aqueous interface.

11:00 AM K3.8/L3.8 
THE NATURE OF THE INTERACTIONS BETWEEN LIPIDS AND ENGINEERED LUNG SURFACTANT PROTEINS. Ka Yee C. Lee, Michael M. Lipp, Joseph A. Zasadzinski, Dept. of Chemical Engineering, University of California, Santa Barbara, CA; Alan J. Waring, MLK/Drew Medical Center and Perinatal Labs, Harbor-UCLA, Los Angeles, CA.

A mixture of lipids and proteins, commonly known as lung surfactant (LS), lines the pulmonary air spaces and facilitates breathing by reducing the surface tension inside the lungs. A deficiency of LS at birth causes neonatal respiratory distress syndrome, a deadly disease which is difficult and costly to treat. Research on replacement therapy has made significant progress; exogenous LS improves oxygenation when administered to premature infants. However, human LS-specific proteins are difficult to obtain, and concerns over immunological responses and viral contaminants make animal proteins sub-optimal. One LS-specific protein (SP-B), important in replacement therapy, has a net positive charge of 8 and is thought to improve the surface activity of LS by preventing anionic lipids from being squeezed-out from the film during exhalation. It appears that the key to the activity of SP-B protein is a balance between cationic and hydrophobic residues in the form of amphipathic a-helical sequences. To test this hypothesis, we have used fluorescence microscopy and isotherm measurements to study the interactions between anionic lipids and 6 different synthetic proteins: (1) full-length SP-B1-78; (2) truncated SP-B1-25; (3) a 21-residue peptide with repeating leucine and lysine residues in the ratio 4:1 (KL4); (4) a SP-B1-25 analog with charged residues replaced by neutral serine residues; and 25-residue homogeneous sequences of (5) hydrophobic leucine residues and (6) cationic lysine residues. We have discovered that the protein sequences containing amphipathic a-helical sequences (1 and 2) lead to an optimal surface activity upon addition to anionic lipid monolayers. Fluorescence microscopy reveals that this activity is due to the ability of these sequences to form a fluid phase network surrounding the condensed lipid phase domains which is able to persist to low surface tensions. This information may help to facilitate the design of simpler and cheaper peptide sequences for LS replacement therapy.

11:15 AM K3.9/L3.9 
A REVERSIBLE, 2- TO 3-DIMENSIONAL PHASE TRANSITION AT HIGH SURFACE PRESSURE IN MONOLAYERS OF PULMONARY SURFACTANT AND SYNTHETIC CONTROL MIXTURES. W. R. Schief, Jr., V. Vogel , Dept of Bioengineering, University of Washington, Seattle, WA; B. M. Discher, Dept of Biochemistry, and S. B. Hall, Dept of Medicine, Oregon Health Sciences University, Portland, OR; D. W. Grainger, Dept of Chemistry, Colorado State University, Fort Collins, CO.

A thin film floating on a water layer lining the lung interior, pulmonary surfactant is a complex mixture of lipids, proteins, and cholesterol which reduces the work of inhalation and stabilizes the small, spherical air sacs (alveoli) against collapse upon exhalation. Our research focuses on revealing the molecular mechanism(s) by which lung surfactant provides the mechanical stabilization, crucial knowledge for the improvement of synthetic lung surfactant administered to premature infants. Employing Langmuir monolayers as a model system, we have combined light scattering-, fluorescence-, and Brewster angle microscopy (BAM) with a high performance teflon ribbon trough to study the phase behaviour of Calf Lung Surfactant Extract (CLSE) and synthetic lipid mixtures as controls over a wide surface pressure range, from 0 mN/m to a plateau in the surface pressure vs. area diagram at 40-50 mN/m. We document the formation at low pressure, and the persistence over a wide pressure range, of condensed monolayer domains in CLSE. Upon compression onto the high pressure plateau, a new 3D phase nucleates in the liquid-expanded region of the monolayer, coexisting with the condensed domains. The new phase appears as bright discs in both fluorescence and BAM. Quantitative analysis of the BAM images indicates that the discs have bilayer thickness and rest above the monolayer. The redistribution of fluorescent probe from the region of liquid-expanded monolayer to the discs suggests that the degree of chain order is lower in the discs. Expansion and recompression of the monolayer shows that this 2D to 3D phase transition is reversible. Experiments reproducing the transition with binary mixtures of DPPC and cholesterol eliminate the possibility that it is an artifact of sample contamination, and further, demonstrate a strong destabilizing impact of cholesterol. Our observations provide new molecular-level information on the high pressure instability of lung surfactant monolayers. We will discuss areas of agreement and disagreement with existing theories of in vivo lung surfactant function.

11:30 AM K3.10/L3.10 
CAMPYLOTAXIS: CURVATURE DRIVEN MEMBRANE-PROTEIN DYNAMICS. L. Mahadevan, Peter T.C. So, and Seth Newburg, Dept. of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA.

The transport and aggregation of membrane proteins are crucial in regulating cellular functions. However, the detailed interaction between membrane forces with the embedded proteins remains poorly understood. Experimental methods used for such studies include fluorescnt photobleaching recovery, fluorescent single particle tracking and gradient optical trap. The consensus emerged from these diverse methods of examination is that membrane protein motion is regulated by multiple mechanisms. Transient confinement by obstacle clusters and tethering to the cytoskeleton, directed motion driven by molecular motors and free random diffusion all play important roles. In addition to these mechanisms, regions of protein aggregation has been observed to correlate with regions of high membrane curvature. In a physical context, protein mobility may be controlled by the regulation of membrane curvature. While a single molecular motor provides precision control of individual receptors, membrane curvature is a global control mechanism whereby many proteins can be simultaneously directed. The goal of this project is to develop theoretical and experimental models to elucidate the motion of biological molecules driven by local regulation of membrane curvature.

11:45 AM K3.11/L3.11 
STRUCTURAL ANALYSIS OF HISTIDINE-TAGGED CAPSID PROTEINS USING METAL-CHELATING LIPID MONOLAYERS. J. Mc Dermott, E. Barklis, Oregon Health Sciences University, Dept. of Microbiology; S. Wilkens, U. Oregon, Institute of Molecular Biology; Y. Rui, D. Thompson, Purdue University, Dept. of Chemistry, West Lafayette, IN.

We have developed a system for analysis of histidine-tagged (his-tagged) retrovirus core (Gag) proteins, assembled onto lipid monolayers consisting of egg L--phosphatidylcholine (EPC) plus the novel synthetic lipid, 1,2-di-O-hexadecyl-sn-glycero-3-N-(5-amino-1-carboxypentyl)iminodiacetic acid (DHGN). AAS indicates that DHGN binds nickel ions. Mixed DHGN/EPC monolayers specifically bind gold conjugates of his-tagged proteins. Two dimensional arrays of HIV and Moloney leukemia virus capsids were crystallized at the air-water interface and analyzed by cryoTEM. Image analysis of these arrays provides 2D capsid structures that are resolved to 20 and 9.5, respectively. The implications of these structures on viral assembly processes will be discussed.

SESSION K4a/L4a: JOINT SESSION: 
BIOMINERALIZATION, BIOMATERIAL DESIGN 
Chair: A. Campbell 
Tuesday Afternoon, December 2, 1997 
Republic B (S)

1:30 PM *K4a.1/L4a.1 
DESIGN PRINCIPLES FOR CONTROL OF CELL FUNCTION BY BIOMATERIALS DESIGN. D.A. Lauffenburger, Dept of Chemical Engineering & Ctr for Biomedical Engineering, MIT, Cambridge, MA.

It is now well-established that improved design of biomaterials requires understanding of how they will interact with cells. Primarily these interactions occur via interactions of materials-associated biochemical ligands with cell surface receptors, generating chemical and physical processes governing cell behavioral responses such as proliferation, adhesion, migration, and differentation. Hence, a central focus of biomaterials design must be aimed toward controlling these ligand/receptor interactions according to quantitative physicochemical principles relating them to cell responses. Although the identities of receptors, their corresponding ligands, and associated signaling pathways involved in these cell responses are becoming increasingly well-known, merely presenting particular ligands on a materials surface is insufficient for controlling desired cell functions. Rather, quantitative aspects of various ligand/receptor interaction properties have a major influence on cell responses to any given ligand, so principles characterizing these effects must be considered prominently. This talk will present an overview of some principles that are being elucidated, with their relevance to biomaterials design.

SESSION K4: BIOMINERALIZATION, BIOMATERIAL DESIGN 
Chairs: A. Campbell and Viola Vogel 
Tuesday Afternoon, December 2, 1997 
Grand Ballroom (S)

2:00 PM *K4.1 
REVERSIBLE HYDROGELS, SELF-ASSEMBLING ARTIFICIAL PROTEINS. Wendy A. Petka, Department of Polymer Science and Engineering, University of Massachusetts, Amherst, MA; James L. Harden, Denis Wirtz, Department of Chemical Engineering, Johns Hopkins University, Baltimore, MD; Kevin P. McGrath, Science and Technology Directorate, U.S. Army Natick RD&E Center, Natick, MA; David A. Tirrell, Department of Polymer Science and Engineering, University of Massachusetts, Amherst, MA.

We discuss the properties of a new class of protein-based biopolymers, obtained using recombinant DNA technology, that contain -helical leucine zipper coiled ends separated by a disordered coil poly(alanylglycine) protein spacer chain. These triblock proteins form environmentally sensitive, thermoreversible hydrogels at protein concentrations above 4% (w/w). The molecular conformation and aggregation behaviors of the proteins in dilute solution are studied as a function of solution pH and temperature using circular dichroism spectroscopy and time-resolved static light scattering. The gel-sol transition and the viscoelastic properties of more concentrated samples are revealed using diffusing-wave spectroscopy, a non-invasive optical rheology technique based on dynamic light scattering. The formation of the protein network is temperature and pH dependent due to the effect of pH and temperature on the intermolecular interactions between helical domains. We also discuss the role of helix architecture in determining the gelation behavior.

2:30 PM K4.2 
ENGINEERED PROTEINS AS RECOGNITION ELEMENTS FOR MATERIALS ASSEMBLY. R. Humbert*, S. Brown**, and Mehmet Sarikaya*; *Materials Science and Engineering, University of Washington, Seattle, WA, **Molecular Cell Biology, University of Copenhagen, Copenhagen, DENMARK.

Structural control of inorganic materials at the nanoscale is a key to the processing of materials with new and improved properties. Biological hard tissues may serve as models for future engineered materials as biocomposites have an intricate nano- and micro-architecture controlled at the molecular level by proteins which have specific interactions with the mineral phase. Combinatorial genetic techniques permit isolation of specific recognition elements for surfaces, including those not realized by natural proteins, in the absence of Apriori prediction of necessary structures. Here we demonstrate the controlled assembly of nanometer-scale gold particles on functionalized spherical and flat surfaces as substrates in aqueous solutions using engineered gold-binding proteins as recognition elements, assembly of proteins to Sat gold surfaces and ultrafine gold particles (120 diam.) to self-assembled protein surfaces were studied by AFM and TEM. The results could have significant implications in tailoring the formation and assembly of ordered structures of metals, functional ceramics, semiconductors, and ferroelectrics in nanotechnology.

2:45 PM K4.3 
PHASE-SELECTIVE SYNTHESIS OF CALCIUM CARBONATE POLYMORPHS. Kenneth M. Doxsee, Department of Chemistry, University of Oregon, Eugene, OR.

The mechanisms whereby Nature controls the morphology and phase of the calcium carbonate biominerals are a matter of ongoing scrutiny. In order to exploit the often dramatic effects of solvent on crystal form, well known in the realm of molecular solids, in the crystallization of extended solids, we have developed the technique of complexation-mediated crystallization, wherein use of solubilizing agents allows the preparation of solid-state materials through simple salt metathesis reactions in nonaqueous solvents. When applied to the crystallization of calcium carbonate through the reaction of calcium chloride with sodium bicarbonate in methanol under ambient conditions, the metastable phase, vaterite, is selectively formed.

3:30 PM *K4.4 
BIOLOGICAL NANOFABRICATION WITH SILICON AND CALCIUM. Daniel E. Morse 1,2,3, Jennifer Cha 1,2,4, Katsuhiko Shimizu 1,2,3, Angela M. Belcher1,2,4, Xueyu Shen 1,2,3, Tilman E. Schaffer 2,5, Galen D. Stucky 2,3,6 and Paul K. Hansma 2,5; 1 Marine Biotechnology Center; 2 Materials Research Laboratory; 3 Department of Molecular, Cellular & Developmental Biology; 4 Department of Chemistry; 5Department of Physics; and 6Department of Materials, University of California, Santa Barbara, CA.

Silicon and calcium yield exquisitely structured high-performance composites produced by biomineralization. While we understand some of the mechanisms by which genetically encoded proteins direct nanofabrication of calcium-based structures, the processes controlling biosynthesis of silicon composites are only now emerging. What are the similarities and differences between the mechanisms used to synthesize and shape composites of these elements? We resolved the activities of 3 families of proteins controlling nanofabrication of the microlaminate calcium carbonate composites of the abalone shell and "flat pearl" Polyanionic proteins determine polymorph selection and atomic lattice orientation by cooperative interactions at the inorganic interface. After a nucleating protein sheet and calcite-specific polyanionic proteins direct nucleation of a "primer" layer of oriented calcite, a genetic switch triggers a sharp transition to synthesis of polyanionic proteins that direct deposition of aragonite on the primer. Microlaminate nacre is organized over macroscopic dimensions by continuous growth of atomically coherent protein-oriented aragonite crystals through pores in a multilayered network of sheets of "matrix" proteins. Stochastic spacing of the stencil-like nanopores in these interlamellar protein sheets determines the lateral offset between successive microlaminae, generating the interdigitating brickwork that gives the material its great strength. The sequence of Lustrin A cloned from this matrix reveals a strikingly repetititve modular structure that helps explain its self-assembly. We have used the purified polyanionic proteins from the abalone shell to sequentially switch crystallographic phase from calcite to aragonite and vice-versa, with stereospecific precision and fidelity, producing multiphase composites with micron-scale domains. The silica spicules of a sponge provide a uniquely tractable model system with which protein control of biosilicification can be analyzed. We have resolved the subunit structure and partial sequences of the integral proteins from a sponge biosilicate, and are using these proteins to direct silica nanofabrication in vitro. Possible mechanisms of action will be discussed.

4:00 PM *K4.5 
MINERALIZATION OF MULTILAYER HYDROGELS AS A MODEL OF MINERALIZATION OF BONE. Paul Calvert, Joelle Frechette and Chad Souvignier, Department of Materials Science and Engineering, University of Arizona, Tucson, AZ.

Extrusion freeform fabrication is a 3-D layerwise writing technique for forming objects directly under the control of a CAD program. This method is one of a family of rapid prototyping methods which include stereolithography, selective laser sintering and fused deposition modeling. This system can be used to build shapes, layer by layer, from hydrogels of agarose, polyacrylamide or other cross-linked water-soluble acrylic polymers. Mineralization can be induced in these gels by building a part with alternating layers of gels containing calcium and carbonate, or phosphate that can be formed into stacks which then mineralize by cross-diffusion. The write head can be conceived of as a cell which delivers the appropriate minerals to a site within a swollen gel matrix. This diffusive mineralization process can thus be compared with bone mineralization. The gel structure controls the morphology of the mineral. The site of mineralization is controlled by osmotic forces which localize most of the mineral in whichever zone has the higher ionic strength. The mineral content, expressed as a fraction of the polymer content, is similar to that of bone but the water content is much higher than in bone. This raises the question of what process drives the water exclusion during bone mineralization.

4:30 PM K4.6 
ASSEMBLY MECHANISM OF A NANOSTRUCTURED BIOLOGICAL HARD TISSUE: A MOLLUSC SHELL. D. W. Frech, T. Graham, R. Humbert, and M. Sarikaya, Materials Science and Eng., University of Washington, Seattle, WA.

Excellent combination of physical properties of biological composites, such as bone, dentin, silicious spicules, and various mollusc shells, is related to their highly ordered hierarchical microstructures that are synthesized from molecular to the nano- and macro-scales under the influence of organic matrices. Some of the beautiful examples are the shell of molluscs. For example, in red abalone, the structure is comparatively simple having two microstructurally different sections; inner nacreous and outer prismatic regions, resulting in a two-tier functionally gradient material for high impact and wear resistance properties. While the prismatic section consists of columnar calcitic crystals, the nacreous section is assembled by aragonitic platelets in submicron layers. New insights have been gained in this microarchitectural material. Our recent invesigation of internal structures and crystallography, by cross-section TEM, and surface structures by SEM and AFM, indicated that prismatic section acts as a substrate for the formation of nacre, by first forming aragonitic nuclei (on parallel ridges) that are ordered pseudohexagonally whose growth follows a diffusion model. Thickening of the nacre section is controlled by the formation of successive aragonite platelets by an anchoring mechanism located within the center of each crystal, originating in calcite. While there is no ìepitaxialî correlation between prismatic and nacreous sections, there is, however, crystallographic correlation among the aragonite platelets within nacre that are hierarchically twinned forming a 3-dimensional tiled structure (a manifestation of pseduhexagonally ordered nuclei). The unique structural build-up (i.e., nanofabrication) and the local nanomechanical properties will be discussed in the context of biomimetic lessons for novel materials design. The research was supported by ARO through URI and AASERT.

4:45 PM K4.7 
MICROSTRUCTURAL DESIGN AND ITS IMPLICATION FOR THE MECHANICAL PROPERTIES OF THE CROSSED-LAMELLAR SHELL: STROMBUS GIGAS. Xiaowei Su, Stephanie Jumel, and A.H. Neuer, Dept of Materials Scinece and Engineering, Case Western Reserve University, Cleveland, OH.

The fracture toughness of the molluskan crossed lamellar shell, Strombus gigas, is three orders of magnitude higher than that of non-biogenic aragonite, the principal component of the shell. The organic components in the shell, and the fact that structure is present at five distinct length scales, both play important roles in determining its properties. Both scannng electron microscopy of partially decalcified samples and transmission electron microscopy of ion-thinned foils show that the EDTA-insoluble organic component is distributed around each third-order lamellae, and at interfaces between first-order and second-order larnellae. The organic components whose volume fraction is 2%, is more prominent at the interfaces between second-order lamellae than that at the other sites. This organic component is a preferred path for crack propagation during catastrophic fracture. During indentation, the crystallites fail when loaded in certain orientations with respect to the orientation of third-order lamellae. The extent of damage also depends on the orientation of the third-order lamellae with respect to the loading axis. 
Damage generation during indentation on the three faces of the long lath-like third-order larnellae was recorded by the depth-sensing indentation method. The load-displacement curves corresponding to the three orientations of the third-order lamellae shows that (1) the depth of the indentation under the same load is the largest when indented on the large face and the smallest when indented on the side faces of the lath-like lamellae, (2) the elastic recovery when indented on side faces is larger than on the top faces and end faces of the lamellae; (3) the permanent deformation is the largest when indented on the end faces, and smallest when indented on the large top faces. The calculated energies and microhardnesses are 1.50x10-6 J and 2.62 GPa, respectively, for indentation on the large top faces 1.36x10-6 J and 2.95 GPa for indentation on the side faces, and I .55xlO-6 J and 3.15 GPa for indentation on the end faces.

SESSION K5: POSTER SESSION 
Tuesday Evening, December 2, 1997 
8:00 P.M. 
Grand Ballroom (S)

K5.1 
COMPARISON OF EXPERIMENTAL WITH THEORETICAL MELTING CURVES OF THE YEAST GENOME AND INDIVIDUAL YEAST CHROMOSOME DENATURATION MAPPING USING THE PROGRAM MELTSIM. J.W. Bizzaro, Kenneth A. Marx and R.D. Blake*, Department of Chemistry, University of Massachusetts, Lowell, MA; Department of Biochemistry, Microbiology and Molecular Biology*, University of Maine, Orono, ME.

The first eukaryotic genome to be completely sequenced as part of the international Human Genome Project is the yeast- S. cerevisiae. Therefore, this genome is of great interest as a eukaryotic model system from a physical chemical, molecular biological, genomics, as well as biomaterials point of view. Using the sequence database for this organism( 12,067,277 bp), we have carried out a simulated melting of yeast DNA using the program MELTSIM and compared it to the experimental high resolution melting behavior of yeast DNA in the following buffer condition: 0.0716 M NaCl, 0.0038 M cacodylic acid, 0.00015 M EDTA, pH 6.7. There is good agreement between the calculated and experimental melts. This demonstration, as well as earlier comparisons for incomplete genomic databases such as D. discoideum (1), validates the utility of MELTSIM to accurately simulate the melting of complex eukaryotic DNAs, whose sequences are becoming increasingly available for study in public databases. We have also simulated the melting of individual yeast chromosomes. In general, there is little variability in melting behavior for the 16 chromosomes. They all undergo largely symmetric melting about similar Tm values, resembling the simulated melting of the overall yeast genome. MELTSIM can also be used to perform denaturation mapping along the individual yeast chromosomes. We illustrate this with representative chromosome denaturation maps which we compare to positional (G+C) base composition and other features along the same yeast chromosome sequence. Comparison of simulated melting of coding and non-coding sequences from individual yeast chromosomes showed a significant and consistent difference in the melting behavior. The coding sequences melt 2-3 degrees higher in temperature, characteristic of a higher GC composition compared to the non-coding sequences. This behavior is similar to what we have observed in other eukaryotic genomes. The ability of MELTSIM to accurately simulate the equilibrium helix-to-coil properties of DNA sequences and their positional denaturation mapping is a valuable tool. It can display a representation of the thermodynamic sequence distribution that has important consequences for the understanding of sequence dependent energetic properties of DNA in their biological context and also for their potential design and use in biomaterials applications. The authors acknowledge support from the Center for Intelligent Biomaterials at UML.

K5.2 
FIRST PRINCIPLES SIMULATIONS OF GLUCOSE IN AQUEOUS SOLUTION. C. Molteni and M. Parrinello, Max Planck Institut fuer Festkoerperforschung, Stuttgart, GERMANY.

Carbohydrates are an important class of biomolecules involved in a variety of processes such as energy storage, structural support, molecular recognition and water control in cold and drought-resistant organisms. There are many problems related to carbohydrates that are still far from being properly understood: a classic example is the behaviour of glucose in aqueous solution. Glucose (whose main structure is a puckered six-membered ring that carries a methyl alcohol and four hydroxyl groups) can be found in two different anomeric forms, alpha and beta, that differ for the orientation (axial or equatorial with respect to the ring) of one of the hydroxyl groups. The experimental ratio alpha:beta in aqueous solution is 36:64; what determines this ratio is still unclear. We have performed first principles molecular dynamics simulations, based on the Car-Parrinello method, of glucose in water, for both the alpha and the beta anomers. By treating in an ab initio way both solute and solvent, we provide an accurate and reliable description of the hydrogen bonds, which determine the water structure around the glucose. Distinct solvation features for the alpha and the beta anomers have been found.

K5.3 
HIGH RESOLUTION TWO-DIMENSIONAL DETECTION OF NCD MOTILITY WITH OPTICAL TWEEZERS. Miriam W. Allersma, Frederick Gittes, Christoph F. Schmidt, Univ of Michigan, Dept of Physics and Biophysics Res Div, Ann Arbor, MI; Michael J. DeCastro, Russell J. Stewart, Univ of Utah, Dept of Bioengineering, Salt Lake City, UT.

The ATP-dependent motility of the kinesin-related mechanoenzyme ncd was observed in a bead motility assay using optical tweezers in combination with a new two-dimensional displacement detection method. In the assay, multiple dimeric fusion proteins of truncated ncd with glutathione-S-transferase (GST-N195) were adsorbed onto 0.5-micron silica beads. The beads walked on reconstituted microtubules (MT) that were immobilized to a glass coverslip, at velocities of approximately 200 nm/sec. Lateral displacements with respect to the MT axis, detected fast enough to avoid low-pass filtering of thermal motions, were consistent with a random exploration of the whole accessible MT surface. If correct, this would mean that ncd does not track a microtubule protofilament. Such behavior would contrast with the high processivity of kinesin and would be similar to that of myosin, dynein and single headed kinesin. The detection technique is based on observing the far-field interference pattern formed by the illuminating laser focus and the light scattered from the trapped dielectric bead. The interference produces shifts in the angular distribution of light within the high numerical aperture cone of light. This translates to lateral light intensity shifts in the back focal plane of the microscope condenser lens collimating the beam, which in turn are very sensitively detected by a quadrant photodiode detector placed at a plane conjugate to the back focal plane of the condenser. The interference effect is described in a scattering approximation for a Gaussian laser focus. The ability to observe the two-dimensional motion, with high temporal and spatial resolution, and in a manner largely independent of position in the microscope field-of-view, is the particular advantage of this detection method. We acknowledge support from the Whitaker Foundation, NSF (BIR-9512699) and the Petroleum Research Foundation.

K5.4 
FORCE TRANSDUCTION IN LISTERIA MOTILITY: IMPLICATIONS FOR POLYMERIZATION-BASED LOCOMOTION. Donald J. Olbris and Judith Herzfeld, Dept. of Chemistry and Keck Institute for Cellular Visualization, Brandeis University, Waltham, MA.

Listeria monocytogenes and a number of other infectious bacteria polymerize a host cell's actin into tails that apparently propel the bacteria through the cytoplasm. The mechanism of force transduction is an open question. By examining the dependence of Listeria motility on viscosity, we calculate that the tail must be attached to the bacterium, or it will drift away in experiments with low viscosity cell-free extracts. Since actin-based Listeria motility is conferred by a single bacterial protein, it is possible that this protein acts as a polymerase. Taking into consideration the relative friction coefficients of the bacteriaís body and the actin tail, we calculate that motor-based bacterial motility would be consistent with observations. We discuss the implications for the mechanisms of polymerization-based motility, including amoeboid cell locomotion.

K5.5 
MOLECULAR DYNAMICS SIMULATION OF THE LS2 ION CHANNEL. Qingfeng Zhong, Qing Jiang, Preston Moore, and Michael L. Klein, Center for Molecular Modeling and Department of Chemistry, Univ of Pennsylvania, Philadelphia, PA; Dennis M. Newns, T.J. Watson Research Center, IBM Corporation, Yorktown Heights, NY.

Recent progress of the molecular dynamics (MD) simulation of the LS2 synthetic ion channel will be discussed. The LS2 ion channel is conceptually simple and resembles the characteristics found in most of the natural ion channels. MD simulations have been performed on the LS2 ion channel for over 6ns in a membrane-mimic enviornment without any artificial constraints. Two distinct metastable states have been identified. The two states have roughly the same energy and each of them lasts for over 2ns. Energetic analysis shows that the hydrophobic interactions between helices and the interactions between hydrophilic groups and pore water are responsible for the helix packing. Moreover, the conformations of hydrophilic groups control the transition between the two states via its interactions with pore water. The pore water shows a different structure than the bulk water. Long wavelength oscillatory collective motions of helical bundle have been indentified. Dynamical properties of helical bundle and pore water inextricably have impact on the transport properties of the ion channel. Based on our simulation, we propose a phenomenological model to describe the static and dynamical properties of helical bundle.

K5.6 
CONFORMATIONAL CHANGES IN BOVINE SERUM ALBUMINE (BSA) AND ITS EFFECT ON TRANSFORMATIONS OF BRUSHITE TO HYDROXYAPATITE: INSITU STUDIES USING FTIR/ATR. Jing Xie, Krishnan Chittur and Clyde Riley, Materials Science Program, Chemistry Dept., Univ. of Alabama in Huntsville, Huntsville, AL.

Cellular response to biological implants is mediated by protein adsorption and subsequent conformation. As physiological solution induces changes in the calcium phosphate ceramic (CPC)coated implant surfaces, it is necessary to monitor such changes and its effect on adsorbing protein. In the present study, FTIR/ATR technique was used to investigate insitu protein adsorption with different CPC surfaces. Calcium phosphate ceramics like brushite (dicalcium phosphate dihydrate, DCPD,CaHPO4.2H2O), transform to more stable forms like hydroxyapatite(HA,Ca5(PO4)3OH), in physiological solution. Adsorption of BSA was studied onto bare germanium and zinc selenide crystals and germanium ATR crystals coated with brushite. Our experiments show that the adsorption of albumin affects the rate of CPC phase transformations and the changing CPC surface had an effect on protein conformation. The FTIR/ATR technique allows the simultaneous in-situ study of protein adsorption and conformational changes while following changes in the composition and structure of CPC ceramics.

K5.7 
ELECTRODEPOSITION OF BRUSHITE AND TRANSFORMATION TO HYDROXYAPATITE: INSITU STUDIES USING FTIR/ATR - CORRELATION WITH X-RAY DIFFRACTION AND SEM/EDS. Mukesh Kumar, Krishnan Chittur, and Clyde Riley, Materials Science Program, Chemistry Dept., Univ. of Alabama in Huntsville, Huntsville, AL.

Calcium phosphates ceramics(CPC) like hydroxyapatite (HA, Ca5(PO4)3OH) are used as coatings on orthopedic implants because they are bioactive and biocompatible. HA is similar to the composition of mineral phase in natural bone. Bioactivity of such CPC phases are probably due to dissolution that leads to locally high concentrations of calcium and phosphate ions. The dissolution properties of HA are strongly dependent on crystallinity and the presence of protein. It has previously been shown that brushite (dicalcium phosphate dihydrate, DCPD, CaHPO4.2H2O), can be used as a coating material because it can dissolve and reprecipitate as HA thereby supplying the necessary calcium and phosphate ions. In the present study, we demonstrate the use of FTIR/ATR to monitor pH induced brushite crystallization in an electrochemical cell and its subsequent transformation to HA by dissolution and reprecipitation in a calcium free physiological solution. Infrared wave guides of germanium were coated with metal film to act as the cathode in the electrochemical cell. Brushite deposition was induced on such wave guides under galvanostatic conditions. The kinetics of deposition of brushite and its subsequent transformation to HA was followed, in-situ, by analysis of changes in phosphate mid-infrared bands. FTIR/ATR data were correlated with information obtained from X-ray diffraction and SEM/EDS analysis.

K5.8 
NANOSCALE BANDING IN OTOLITHS FROM ARCTIC CHARR AND PIKE: A TEM STUDY. A. Meldrum Oak Ridge National Laboratory, Solid State Division, Oak Ridge, TN; N. Halden, University of Manitoba, Department of Geological Sciences, Winnipeg, Manitoba, CANADA.

Transmission electron microscopy of otoliths from the inner ear of arctic charr and pike has revealed the presence of fine banding on the scale of several nanometers. The thickness of the bands was observed to vary in different portions of the sample, and some areas were not banded. EDS analysis could not detect chemical differences within the bands, but electron diffraction showed that the crystallographic orientation of the bands are related by a low-angle offset. Previously, micron-sized banding in otoliths from arctic charr was observed by SEM and was attributed to seasonal migrations from a fresh water to a marine environment. We hypothesize that the fine-scale banding observed in this study may represent a daily variation. Electron diffraction from the pike samples shows that the material is composed of CaCO3 having the vaterite (as opposed to aragonite) structure and hydrous CaCO3 was also observed. The large-scale banding previously identified by SEM was not observed in the TEM, despite attempts to intersect the boundaries of the micron-sized layers. The interaction of the electron beam with the sample material was investigated by conducting several electron-irradiation experiments. The electron beam was observed to interact strongly with the samples, and caused the precipitation of cubic CaO from the calcium carbonate matrix. Bright-field imaging showed the development of fine-grained ( 5 nm) randomly-oriented CaO crystallites which accumulated with increasing electron dose. These initial results suggest that the precipitation of CaO is not driven by electron-beam heating. Previously, a similar phase-change phenomenon has been observed in hydroxyapatite from dental enamel. Other Ca-bearing biominerals may therefore be expected to be similarly sensitive to electron irradiation.

K5.9 
MICROSCALE RADIATIVE EFFECTS IN COMPLEX MICROSTRUCTURES OF IRIDESCENT BUTTERFLY WING SCALES. Haruna Tada, Seth E. Mann, Ioannis N. Miaoulis and Peter Y.l Wong, Thermal Analysis of Materials Processing Laboratory, Tufts University, Medford, MA.

Multilayer thin-film structures found on butterfly wing scales produce iridescence due to thin-film interference, diffraction, and non-planar specular reflection. The optical phenomena depend heavily on the complex microstructures of the scales. The scales resemble a flattened sack, gathered at one side, with a series of longitudinal ridges rising from the top layer. In many species, the space between top and bottom layer, or the ridges themselves, consist of alternating layers of air and chitin thin-films, each approximately 0.01 m thick. The spectral reflectivity of butterfly wing scales are affected by the thin-film structure, as well as by the angles of incidence and view. Two species of iridescent butterflies, Morpho menelaus and Papilio blumei , were studied on macroscale and microscale level to determine the effects of the cellular microstructures on the spectral reflectivity of the butterfly wing scales at various angles of incidence and view. The thin-films on butterfly wings are multifunctional; their functions include camouflage, display, and thermoregulation. In addition, their structure is similar to human-made microelectronics, and the understanding of the optical properties will not only result in an understanding of butterfly behavior, but also of the radiative effects during microelectronic processing. On the macroscale level, the results show colors from blue-green to violet and yellow-green to violet for M.menelaus and P. blumei , respectively, at various angles of incidence and view. This is produced by diffraction for M. menelaus , and by non-planar specular reflection for P. blumei . On the microscale level, models of the thin-film structures are developed to numerically determine the spectral reflectivity of butterfly wings. The spectral reflectivities at normal incidence are measured experimentally using the microscale reflectance spectrometer. The results show that at normal incidence, the wing scales have high reflectivity only at very selective wavelengths. At other wavelengths, the reflectivity is low, allowing the butterfly to efficiently absorb solar radiation.

K5.10 
SELF-ASSEMBLY OF TYROCIDINES IN NANOTUBULAR STRUCTURES. M. Thies, H.H. Paradies*, * Biotechnology & Physical Chemistry, Markische Fachhochschule, Iserlohn, GERMANY & University of Paderborn, Fachbereich Chemie& Chemietechnik, Paderborn, GERMANY.

Hollow tubular structures of tyrocidine, a cyclic octapeptide with 4 amino acid residues in the D-configuration and 4 in the L-configuration, were produced in aqueous solutions in the presence of ethanol and various amounts of urea. [1]. SAXS, SANS as well as static & dynamic light scattering studies confirm an arrangement of the cyclic monomers in the self-assembled material in form of a tubular structure which can grow into larger tubular structures of nanometers with internal diameters of 0.3 nm. These tubes are open ended, with uniform shape and constant internal diameter. The growth into tubular structures can be regulated through the addition of urea in the aqueous-ethanol solution. FT-IR analysis of the tubular aggregates reveal similarities with those structures reported for gramicidin A, a linear peptide, known to form dimeric -helical transmembrane ion channel structures [2].Tyrocidines, however, forms basically a trimeric & cyclic structure before growing into large tubules.

K5.11 
QUANTUM ELECTRODYNAMICS CALCULATION OF LONG-RANGE INTERACTION OF BIOPOLYMERS IN INTRACELLULAR DISPERSION MEDIUM. Anatoliy Pinchuk, Vladimir Vysotskii, Kyiv Shevchenko University, Radiophysical Dept, Kyiv, UKRAINE.

The physical mechanism of long - range interaction between nucleotides in intracellular liquid medium has been theoretically studied. The analysis of the forces , acting in liquid medium between the nucleotides , situated on the opposite ends of a broken DNA's helix was performed on basis of well known in physics long - range Van der Waals force with taking into account the frequently -dispersion structure of real interaction biological objects. As this takes place , a most intriguing difficulty in calculation of the sign and value of the resulting force of interaction arises from the indispensability of taking into account of dispersion characteristics of the nucleotides and the aqueous salt - containing medium within the whole spectral range up to X - ray frequencies .It was shown that accounting of infrared and ultraviolet resonance's in dielectric functions of DNA's and intracellular liquid adds up to less then 10 percents in the energy interaction and thus the fundamental contribution in the interaction is presented by the group of resonances with frequencies of X - ray range. The computer simulation of energy interaction has shown that during the interaction between thiamin - guanine , adenine -guanine and cytosine -guanine there exists a potential barrier over a distance of about 10 - 20 angstroms , which prevents the enzyme selfrepairing in DNA's after a double damage . All the remaining pairs of nucleotides have no such a barrier . It was shown that with the increasing of the temperature and diminishing of the viscosity of intracellular medium the barrier vanishes and DNA's undergoes a complete selfrepairing . The quantum -electrodynamics mechanism under consideration complements the conventional scheme of the enzymatic repairing of DNA's damages and gives the methods for improving the self - appearing of DNA's breaks.

K5.12 
PROTEIN-INDUCED FOLDING PROTRUSIONS ELIMINATE IRREVERSIBLE LIPID SQUEEZE-OUT IN MODEL LUNG SURFACTANT MONOLAYERS. Michael M. Lipp, Ka Yee C. Lee, Joseph A. Zasadzinski, University of California at Santa Barbara, Dept. of Chemical Engineering, Santa Barbara, CA; Alan J. Waring, University of California at Los Angeles, Dept. of Pediatrics, Los Angeles, CA.

Lung surfactant (LS) is a mixture of lipids and proteins which facilitates respiration by reducing the alveolar surface tension. To function properly, LS must: (1) adsorb rapidly to the alveolar air/water interface, (2) form a monolayer which can achieve low surface tensions upon compression, and (3) respread rapidly and reversibly upon monolayer collapse. The lipids found in LS can be classified into two categories: saturated phospholipids capable of forming monolayers allowing for low surface tensions, and unsaturated and anionic fluidizing lipids which can adsorb and respread well but cannot achieve low surface tensions. This dichotomy in behavior has led to the hypothesis that the fluidizing lipids in LS are squeezed-out of the monolayer to leave it enriched in saturated lipid. However, we have discovered that SP-B, a cationic protein found in LS, interacts selectively with fluidizing anionic LS lipids to allow both to remain in monolayers containing saturated lipids up to collapse at low surface tensions. Using fluorescence microscopy, we have observed that SP-B protein partitions preferentially into fluid phases in mixed monolayers during phase transitions, forming a network surrounding the condensed phases of saturated lipid. In the presence of protein, this fluid network is seen to be retained in the monolayer up to collapse (in contrast to the irreversible loss of the fluid phase into the subphase seen prior to collapse in the absence of protein). Upon collapse, the monolayer folds and forms protrusions extending into the subphase. The protrusions appear to retain the same composition of the original monolayer, and also reincorporate rapidly and reversibly into the monolayer upon expansion. This synergistic interaction between the fluidizing LS lipids and SP-B protein may explain how the complete LS mixture can both form low surface tension monolayers and adsorb and respread rapidly.

SESSION K6: MEMBRANES AND VESICLES 
Chairs: Evan Evans and Michael Schick 
Wednesday Morning, December 3, 1997 
Grand Ballroon (S)

8:30 AM *K6.1 
ORDER IN MEMBRANES AND MONOLAYERS. David R. Nelson, Lyman Laboratory of Physics, Harvard University, Cambridge, MA.

Dislocations and disclinations, the basic ingredients of a theory of melting in flat two dimensional monolayers, play an important role in fluctuating membranes as well. We review some basic facts about defects in membranes and monolayers, and discuss their influence on the size distribution of spherical vesicles composed of lipid bilayers.[1] There is a well developed theory of equilibrium size distributions for dilute suspensions of vesicles with liquid-like internal order.[2] However, all such vesicles should enter a hexatic phase at sufficiently low temperatures, provided the solvent does not freeze first. We give arguments for an unusual distribution close to the liquid-to-hexatic transition, with a diverging average vesicle size which scales with the hexatic correlation length.

9:00 AM K6.2 
SPONTANEOUS VESICLES: THEORETICAL PREDICTION, EXPERIMENTAL VERIFICATION, AND TRIGGERED DRUG DELIVERY APPLICATIONS. D. H. Thompson, O. V. Gerasimov, R. Haynes, J. Boomer, & I. Szleifer, Department of Chemistry, Purdue University, West Lafayette, IN.

Single chain mean field theory has been used to predict the conditions under which thermodynamically-stable, spontaneous vesicles will form. These predictions have been verified by experimental results using egg phosphatidylcholine and dipalmitoyl-sn-phosphatidylcholine as vesicle-forming lipid containing minor proportions of PEG-modified phospholipids. The average particle size, polydispersity, and entrapped volume of these spontaneous vesicles is similar to those formed by extrusion techniques using 0.1 mm track etch polycarbonate membrane filters. Theoretical predictions, experimental results (light scattering, electron microscopy, and 31P NMR), and drug delivery applications of these stable vesicles will be discussed.

9:15 AM K6.3 
STATIC AND OSCILLATORY INSTABILITY OF A LIPID MEMBRANE INDUCED BY INTERACTION WITH A NONEQUILBRIUM DIFFUSION FIELD. Isabelle Durand and Chaouqi Misbah, Laboratoire de Spectrometrie Physique, Universite Joseph Fourier, Grenoble Saint Martin d'Heres, FRANCE.

A model for interaction of a lipid membrane and a nonequilibrium diffusion field is developed. The membrane is coupled to a particle resevoir set at some distance, and permanentlty maintained at a constant concentration. The concentration beneath is kept constant via channels. The molecules diffuse in the bulk, attach to ,or detach from, the membrane, diffuse along it, and ultimately cross it. The membrane motion is limited both by diffusion and hydrodynamics flow. Under wide circumstances we find that the membrane undergoes a morphological instability. The instability can be static or oscillatory in time (Hopf bifurcation). This feature may be evoked to interpret some aspect of vesicle budding in the Golgi apparatus.

9:30 AM K6.4 
SPONTANEOUS CURVATURE OF BILAYER MEMBRANES. Hans-Guenther Doebereiner, Max-Planck-Institute for Colloids & Interfaces, Teltow, Germany.

The morphology of fluid lipid vesicles is governed by the bending elasticity of their membrane which is characterized by the bending modulus and the spontaneous curvature of the bilayer. Despite its fundamental role, the latter quantity has received very little experimental attention so far. Here, I present a recently developed technique for the measurement of the spontaneous curvature using quantitative phase contrast microscopy. For the first time, we are able to monitor the spontaneous curvature as a function of transversal membrane asymmetry. I will show that the tendency of a closed membrane to bend in mechanical equilibrium has two different physical origins corresponding to the interactions of the two monolayers with the adjacent solutions and the bilayer architecture respectively. Further, I present results on specific systems including one and two component unilamellar vesicles in asymmetric aqueous sugar and polymer solutions. Especially, we demonstrate the effect of membrane anchored poly (ethylene glycol) on the shape of vesicles.

9:45 AM K6.5 
ELECTRO-OPTICS OF UNILAMELLAR VESICLES. Niloofar Asgharian, Z. A. Schelly, University of Texas at Arlington, Center for Colloidal and Interfacial Dynamics, Department of Chemistry and Biochemistry, Arlington, TX.

Electric field-induced transient birefringence and light scattering are reported for aqueous suspensions of synthetic unilamellar bilayer vesicles, prepared from the lipid dioleoylphosphatidylcholine (DOPC). The multiexponential birefringence relaxations on the microsecond timescale are interpreted in terms of the field-induced deformation, reorientation, and linear chain formation (and possible electroporation) of the vesicles. Above certain threshold values of the electric field strength and vesicle concentration, field-induced light scattering occurs concomitant with the birefringence. The corresponding transient signals corroborate the linear chain formation of the induced dipolar vesicles. The birefringence relaxation data are used for calculating the bending elasticity modulus of the bilayer.

10:30 AM *K6.6 
ENGINEERING ENTROPY: BUILDING ORDER WITH DISORDER. Seth Fraden, Complex Fluids Group, Department of Physics, Brandeis University, Waltham, MA.

Hard particles are impenetrable objects with no other interaction potential. Although lacking interparticle attractions, hard particle fluids exhibit rich phase behavior that is controlled solely by the shape and concentration of the particles. We describe our work on liquid crystal phases in virus suspensions, with emphasis on how the phase behavior can be understood as arising entirely from entropic considerations. We report experimental studies, computer simulations, and statistical mechanical models of mixtures of colloidal rods and spheres. Surprisingly, for special ratios of the rod-length to sphere-diameter, phase separation can occur on a microscopic scale, leading to alternating layers of spheres on rods in one case, or to columns of spheres assembling into lattices interspersed with rods.

11:00 AM K6.7 
ADHESION OF LIPID MEMBRANES MEDIATED BY ELECTROSTATIC AND SPECIFIC INTERACTIONS. Christian W. Maier, Almuth Behrisch, Annette Kloboucek, Rudolf Merkel, Technical University Munich, Dept of Physics, Garching, GERMANY.

Adhesion of biological cells is caused by specific bonds between adhesion molecules as well as universal interactions. In order to reduce the complexity of the system, we used unilamellar giant vesicles as simple models for real cells. Energies of adhesion were determined by the micropipet technique (E. Evans, 1980, Biophys. J., vol. 31, p. 425). We investigated membrane adhesion created by universal interactions. The strength of adhesion was mediated by addition of charged lipids to the neutral host membranes. Both anionic and cationic lipids were used. The charge densities of the adherent bilayers were systematically varied. Surprisingly, negative charges seemed to be more effective in regulating adhesion of membranes than positive ones. In a separate series of experiments, adhesion was caused by incorporation of a specific cell adhesion molecule, contact site A (csA) from Dictyostelium discoideum. In this case, the characteristics of specific, quasi irreversible bonding were found. The measured adhesion energy is determined by the number of bonds and increased with time. This increase is caused by unbound molecules diffusing into the adhesion zone where bonding takes place. Under low salt conditions, compression of the receptors in the adhesion zone was reversible. In contrast to this, we found irreversible aggregation of adhesion molecules after compression under physiological salt conditions (150 mM sodium chloride).

11:15 AM K6.8 
MODEL SYSTEM FOR LOCAL ADHESION SITES ON BIO-MEMBRANES. Rebecca Menes, Materials Research Laboratory, University of California Santa Barbara; Samuel A. Safran, Weizmann Institute of Science, Rehovot, ISRAEL.

Recent Experiments1 on bound bilayer membranes show a strong local unbinding when the membranes are locally pinched together by optical tweezers. These experiments are model systems for the study of the behavior of sticking sites, such as the molecular stickers involved in cell adhesion. 
We study, theoretically, the properties of such a sticking site by introducing a non-linear model which can accommodate for the strong perturbation of the initially flat bilayers. We find that the membrane profiles near such a sticker strongly overshoot. The system shows several scaling regimes, where the overshoot size scales as an inverse power law of the pinching strength. This non-linear scaling behavior affects the interactions between the stickers, and leads to interesting aggregation properties which are important in understanding the initial stages of cellular adhesion.

11:30 AM K6.9 
WETTING AND CAPILLARY CONDENSATION AS MEANS OF PROTEIN ORGANIZATION IN MEMBRANES. Tamir Gil, Mads C. Sabra, John Hjort Ipsen, and Ole G. Mouritsen, Dept of Chemistry, The Technical Univ of Denmark, Lyngby, DENMARK; Carlos F. Tejero, Facultad de Ciencias Físicas, Univ Complutense de Madrid, Madrid, SPAIN.

Wetting and capillary condensation are thermodynamic phenomena in which the special affinity of interfaces to a thermodynamic phase, relative to the stable bulk phase, leads to the stabilization of a wetting phase at the interfaces. Wetting and capillary condensation are here proposed as mechanisms that in membranes may serve to induce special lipid phases in between integral membrane proteins leading to a long-range lipid-mediated joining forces acting between the proteins and hence providing as means of protein organization. The consequences of wetting in terms of protein aggregation and protein clustering is illustrated within a concrete calculation on a microscopic model of lipid-protein interactions that accounts for the lipid-bilayer phase equilibria and direct lipid-protein interactions governed by hydrophobic matching between the lipid-bilayer hydrophobic thickness and the length of the hydrophobic membrane domain. A simple analytic method for calculating the phase diagram of protein gas, liquid and solid is developed on the basis of a phenomenological capillary potential. The theoretical results are expected to be relevant for optimizing the experimental conditions required for forming protein aggregates and regular protein arrays in membranes.

11:45 AM K6.10 
SMALL-ANGLE X-RAY SCATTERING STUDIES OF PHOSPHOLIPASE A2 ACTIVITY IN MULTILAMELLAR LIPID BILAYERS. Jesper lemmich, Condensed Matter Physics and Chemistry Department, Riso National Laboratory, DENMARK; Frank Richter, EMBL Outstation at DESY, Hamburg, GERMANY; Thomas Hö nger, Dept. of Chemistry, The Technical University of Denmark, Lyngby, DENMARK.

We have studied the time-evolution of the small-angle X-ray scattering of multilamellar DPPC bilayers upon addition of phospholipase A2 in the region close to the main phase transition, where the enzyme activity undergoes dramatic changes. A comparative study has been performed on the ternary system: DPPC/1-lyso-PPC/palmitic acid in order to mimic a system, where a certain degree of the lipid has been hydrolysed due to the enzyme. In this system a very rich phase behavior is observed in the temperature region close to the main phase transition of the pure lipid.

SESSION K7: SUPPORTED BILAYERS 
Chairs: Anne L. Plant and Viola Vogel 
Wednesday Afternoon, December 3, 1997 
Grand Ballroon (S)

2:00 PM K7.2 
HYBRID BILAYER MEMBRANES IN AIR AND WATER: INFRARED SPECTROSCOPYAND NEUTRON REFLECTIVITY STUDIES. Curtis W. Meuse, Susan Krueger, Anne L. Plant, Biotechnology Division and NIST Center for Neutron Research, National Institute of Standards and Technology, Gaithersburg, MD.

Phospholipid/alkanethiol hybrid bilayer membranes can be prepared in air by horizontal transfer of phospholipid from the air/water interface onto a self-assembled alkanethiol monolayer on gold. We have characterized the resulting hybrid bilayer membrane in air using atomic force microscopy, spectroscopic ellipsometry, and reflection-absorption infrared spectroscopy. These analyses indicate that the phospholipid added is one monolayer thick, is continuous, and exhibits molecular order which is similar to that observed for phospholipid/phospholipid model membranes. The hybrid bilayer prepared in air has also been re-introduced to water and characterized using neutron reflectivity. Neutron scattering from these samples was collected out to a wave vector transfer of 0.25 ‰ -1, and provided a sensitivity to changes in total layer thickness on the order of 1-2 ‰. The data confirm that the acyl chain region of the phospholipid layer is consistent with that observed for phospholipid-phospholipid bilayers, but suggest greater hydration of the phospholipid headgroups of HBMs than has been reported in studies of lipid multilayers.

2:15 PM K7.3 
CHARACTERIZATION OF SELF-ASSEMBLED MONOLAYERS (SAMS) OF ALKYLATED 3,6,9,12,15,18-HEXAOXATHIOLS ON GOLD (AU) AND HYBRIDE BILAYER MEMBRANES DERIVED THEREFROM. David J. Vanderah, Curtis W. Meuse, Scott Glazier, Anne L. Plant, Biotechnology Div., Natl. Inst. of Standards & Technology, Gaithersburg, MD; Vitalii Silin, Dept. of Chemistry, Georgetown Univ., Washington, DC; Luan Nguyen, Dept. of Chemistry, Univ. of California-Irvine, Irvine, CA.

Novel linear thiols containing a 1-thiahexa(ethylene oxide) [THEO] moiety, i.e., HS(CH2CH2O)6R, where R = linear alkyl hydrocarbon, were prepared for evaluation as useful supported hybride bilayer membrane (HBM) materials. SAMs of these compounds were prepared and studied by relection absorption infrared spectroscopy (RAIRS), spectroscopic ellipsometry, electrochemistry, and contact angle measurements. These data indicate that the SAMs are hydrophobic, as expected, and well-ordered possessing a 7/2helix for the THEO moiety oriented normal to the substrate with the alkyl tail in essentially an all trans extended configuration canted at 30 5o to the normal. Kinetics of SAM and bilayer formation were determined by surface plasmon resonance (SPR). Structural characterization of HBMs of these compounds will be presented.

2:30 PM K7.4 
SYNTHESIS AND CHARACTERIZATION OF SUPPORTED BIOACTIVE LIPID MEMBRANES: MODEL SUBSTRATES FOR TISSUE ENGINEERING. Theodore M. Winger, Elliot L. Chaikof, School of Chemical Engineering, Georgia Institute of Technology and Dept. of Surgery, Emory University.

We have postulated that cytomimetic surfaces may provide a useful strategy for biomaterial facilitated tissue engineering. To date, model studies have focused on the synthesis and characterization of substrate-supported self-assembled monolayers (SAM) of zwitterionic phospholipids and phospholipidated bioactive peptide molecules. SAMs have been formed by fusion of extruded vesicles (200nm in diameter) onto smooth hydrophobic borosilicate glass (Ra < 1.5‰). SAMs of dipalmitoyl- (DPPC), dimyristoyl- (DMPC), and dilaurylphosphatidylcholine (DLPC) were constructed in presence or absence of cholesterol. In absence of cholesterol, the topography of the self-assembled system proved to be very smooth by atomic force microscopy (AFM), with an Ra of less than 1.5‰/100 sq. µm. AFM depth measurements of detected small holes yielded values of 22‰ for DPPC and 13‰ for DLPC, respectively, which agrees with predicted monolayer thickness. Further support of monolayer formation by vesicle fusion was provided by experiments implementing radiotitration of C14-labeled DPPC as a biomembrane probe. In presence of 33% cholesterol monolayer holes were not detected, confirming the role of cholesterol as a biomembranes plasticizer. However, SAM roughness did increase (2-5‰) due to the presence of partially fused liposomes. Current studies are directed at characterizing film structure-property relationships including molecular mixing behavior, as well as biomembrane mechanical and chemical stability.

2:45 PM K7.5 
STRUCTURE AND INTERACTION FORCES BETWEEN SOFT-SUPPORTED PLANAR LIPID BIOMEMBRANES. Joyce Wong, Chad Park, Markus Seitz, and Jacob Israelachvili, Dept of Chemical Engineering, University of California at Santa Barbara, CA.

Lipid bilayer systems have been used extensively to study the structure and function of biomembranes including recognition adhesion and fusion. Phospholipid bilayers supported on flat solid substrates have been used to model interactions between membranes, but have several drawbacks: (i) the membrane fluidity is decreased since it is constrained by the substrate surface; and (ii) it is difficult to insert transmembrane proteins into the membrane. 
Our approach is to create biomembranes supported on a polymer cushion which can swell and act as a deformable and mobile substrate, thus resembling the cytoskeletal support in actual cells. We adsorbed a highly branched cationic polymer (polyethyleneimine, PEI) onto mica substrates, and phospholipid bilayers were formed via adsorption of small unilamellar vesicles or in combination with Langmuir-Blodgett techniques. We will focus on investigating the structure of these assemblies using the surface forces apparatus, atomic force microscopy, fluorescence microscopy, and neutron reflectivity. Together, these studies give insight into how properties such as adhesion and fusion between biomembranes is modulated when the biomembrane is ``free'' or on a soft support, rather than bound to a surface.

SESSION K8: CELLS 
Chairs: Anne L. Plant and Viola Vogel 
Wednesday Afternoon, December 3, 1997 
Grand Ballroon (S)

3:30 PM *K8.1 
HEPATIC TISSUE ENGINEERING USING MICROFABRICATION. Sangeeta N. Bhatia, Martin L. Yarmush and Mehmet Toner, Center for Engineering in Medicine, and Surgical Services, Massachusetts General Hospital, Harvard Medical School, and Shriners Burns Institute, Boston, MA.

The repair or replacement of damaged tissues using in vitro strategies has focused on manipulation of the cell environment by modulation of cell-extracellular matrix interactions, cell-cell interactions, or soluble stimuli. Development of functional tissue substitutes through ``tissue engineering'' has been facilitated by the ability to control each of these environmental influences; however, in co-culture systems with two or more cell types, cell-cell interactions have been difficult to manipulate precisely. The ability to spatially control cells at the single cell level using micropatterning would allow the precise manipulation of cell-cell interactions of interest. We have developed an adaptable method for generating two-dimensional, anisotropic model surfaces capable of organizing two different cell types in discrete spatial locations. We have chosen a primary rat hepatocyte/3T3 fibroblast cell system due to its potential clinical significance in bioartificial liver design and also based on widely reported interactions observed in this co-culture model. We have used photolithography to pattern biomolecules (collagen I) on glass which mediates cell adhesion of the first cell type, hepatocytes, followed by non-specific, serum-mediated attachment of fibroblasts to the remaining unmodified areas. This co-culture technique allowed the manipulation of the initial cellular microenvironment without variation of cell number. Specifically, we were able to control the level of homotypic and heterotypic interactions in co-cultures over a wide range. Modulation of initial cell-cell interaction was found to have significant effects on liver-specific markers of metabolic, synthetic, and excretory function. In particular, 2 to 3-fold variations in steady-state levels of representative hepatocellular functions were achieved from identical numbers of cells. The mechanisms by which hepatocytes and fibroblasts interact to produce a differentiated hepatocyte phenotype were also investigated. Variations in bulk tissue function were due to spatial heterogeneity in the pattern of induction of hepatocyte differentiation within a hepatocyte population due to interaction with mesenchymal cells. We found that hepatocytes adjacent to the heterotypic expressed increased levels of intracellular albumin (a marker of hepatic synthetic function), whereas, hepatocytes far from the heterotypic interface contained undetectable levels of albumin. Although the actual molecular basis of this signaling was not identified, our experimental results indicated that the source of the observed induction pattern was a tightly cell-associated fibroblast product. This general approach was potential applications in many areas of tissue engineering, implantation biology, and developmental biology, both in the arena of basic science and in the development of cellular therapeutics.

4:00 PM K8.2 
CELL SORTING IS A PHASE ORDERING PROCESS. Arpita Upadhyaya, Sridhar Raghavachari, Richard C. Raphael, James A. Glazier, Dept. of Physics, Univ. of Notre Dame, Notre Dame, IN; D. Beysens, CEA Grenoble, FRANCE; Gabor Forgacs, Dept. of Physics and Biology, Clarkson Univ., Potsdam, NY.

Sorting of cells in randomly mixed aggregates of two cell types is very similar to the phase separation of two immiscible liquids. We study the kinetics and morphology of cell sorting from embryonic tissues (chick pigmented and neural retinal cells). We find that, in accord with recent theory and experiments in fluids under zero-gravity, the tissue evolution results from coalescence of droplets that rearrange due to an effective surface tension. For large volume fractions, the drops continuously coalesce to form an interconnected pattern, with typical wavelength growing like t. At lower volume fractions we expect a t1/3 growth law. We are currently doing experiments to measure the growth law for the characteristic cluster size at low volume fractions and determine the critical value for the transition between the two growth laws. We are also determining colaescence times and diffusion constants of single pigmented cells and small cell clusters in aggregates of neural cells. The important quantity is the ratio of the diffusion time to the coalescence time. These measurements will specify parameters for comparison with a hydrodynamic model for droplet coarsening.

4:15 PM K8.3 
THERMODYNAMIC PRINCIPLES IN THE SORTING AND PATTERNING OF CELLS ON SURFACES. Mark J. Powers1,2, Cathryn A. Sundback1,2, Stephen S. Kim2, Mark Benvenuto2, Joseph P. Vacanti22%%2, Linda G. Griffith^1,^1Massachusetts Institute of Technology, Department of Chemical Engineering and Center for Biomedical Engineering, Cambridge, MA;^2Harvard Medical School and Children's Hospital, Department of Surgical Research, Boston, MA.

The ability to control the formation of tissue-like structures from dissociated cells is a central goal in tissue engineering research. For many applications this involves the formation of structures comprised of multiple cell types which sort to exhibit specific cellular arrangements or architectural patterns. We have previously demonstrated that the morphogenesis of cellular structures consisting of a single cell type (hepatocytes) is dependent on the biophysical features of the substratum. This analysis is now extended in an attempt to understand the ability of substrata to guide the formation of tissues composed of multiple cell types. It has been shown by other researchers that cells are capable of sorting from one another using principles of free energy minimization. That is, differences in the number of cell-cell adhesion receptors between cell types provides disparity in the relative ìsurface tensionsî of these cells. When these cell types are then cocultured in suspension they sort from one another based strictly on these thermodynamic differences. We now utilize these principles as they apply to cocultured cells on surfaces. In addition to the thermodynamic principles discussed above, the substratum surface tension (i.e., cell-substratum adhesion) must be considered in this analysis. Using a model system of liver tissue cells (hepatocytes and endothelial cells) we have demonstrated that a wide range of architectural patterns are both predicted and achieved simply by altering the adhesive properties of the substratum. Cells sort from one another as monolayers, multilayers, or speroids depending on whether the substratum is of high, intermediate or low adhesivity respectively. These results suggest that in many cases the existing thermodynamic properties of the cells or the substratum can be exploited to achieve desired cellular structures and serve to provide a rational basis for the design and assessment of materials for use in tissue engineering applications.

4:30 PM K8.4 
EFFECT OF CELL AGE AND ULTRASOUND ON THE PERMEABILITY OF CELL WALLS OF GRAM-NEGATIVE BACTERIA. Natalya Rapoport, University of Utah, Dept of Bioengineering, Salt Lake City, UT; William G. Pitt, Brigham Young University, Provo, UT.

Various factors affecting the permeability of cell walls of gram-negative bacteria have been elucidated using the spin-labeling EPR method. The penetration and distribution of a lipophilic spin probe, 16-doxylstearic acid (16-DS) was used to characterize the permeability of phopholipid membranes, whereas the bioreduction rate of hydrophilic probes characterized the permeability of porin channels. Structural heterogeneity of phospholipid membranes was observed, with the presence of structurally stronger and weaker sites characterized by different susceptibility to the surfactant treatment. EPR spectra indicated that ultrasound enhanced the penetration of the lipophilic spin probe into the structurally stronger membrane sites. The effect of ultrasound on cell membranes was transient. The initial membrane permeability was restored upon termination of ultrasound treatment. Ultrasound treatment did not affect the permeability of porin channels. EPR spectra of 16-DS indicated that membrane packing at the sites of probe localization was less dense for younger cells, which might be related to the observation that younger cells are more sensitive to antibacterial agents.

SESSION K9: CELL ADHESION, MOTILITY, TRANSPORT 
Thursday Morning, December 4, 1997 
Grand Ballroon (S)

8:30 AM *K9.1 
GROWTH OF CELLS ON MICROPATTERNED SURFACES. G. Whitesides, Harvard University, Department of Chemistry, Cambridge, MA.

Abstract not available

9:00 AM K9.2 
POTENTIAL DEPENDENT ENDOTHELIAL CELL ADHESION, GROWTH AND CYTOSKELETAL REARRANGEMENTS. Tiean Zhou, Susan J. Braunhut, Diane Medeiros and Kenneth A. Marx*, Departments of Chemistry* and Biological Sciences, University of Massachusetts, Lowell, MA.

Normal endothelial cells (ECs), lining the blood vessels, function to regulate blood flow, create a selective barrier to the interstitial space and interact with circulating inflammatory cells during injury or infection. In principle, these functions of ECs are influenced by their interaction with the underlying extracellular matrix (ECM) which may produce a piezoelectric potential under hemodynamic stress conditions. The possibility that this environmental influence of the ECM biomaterial matrix may affect the EC metabolic state and functions in vivo prompted us to study the response of cultured ECs on indium-tin oxide (ITO) glass subjected initially to varying DC surface potentials (-0.3 to +1.0 V wrt Ag/AgCl). Following 1 hr of exposure to a constant surface potential, we measured, relative to controls, cellular viability, growth rate and changes in actin microfilament organization in ECs over the subsequent 9 days in culture. The growth rate of ECs was inhibited by positive surface potential and differences could be detected as early as three days. We also observed a potential dependent cellular shape change and actin microfilament rearrangement at higher potentials within three days of treatment. ECs changed in average cell surface area and assumed a polygonal cell shape in response to treatment. Using NBD-phalloidin stain for actin and fluorescence microscopy, microfilaments were observed to redistribute to the periphery of the cell, indicative of stress. We intend to pursue direct studies of the ECM matrix (fibronectin, laminin and collagens) to determine its potential dependent and piezoelectric biomaterials properties. This may illuminate ECMs potential role in vivo in signal transduction to the anchored ECs.

9:15 AM K9.3 
COLLECTIVE EFFECTS IN SURFACE-ATTACHED CELL MIGRATION. Andras Czirok, Tamas Vicsek, Eotvos Univ, Dept of Atomic Physics, Budapest, HUNGARY; Katalin Schlett, Emilia Madarasz, Eotvos Univ, Dept of Comparative Physiology, Budapest, HUNGARY.

The trajectories of various surface attached cells have been analyzed with computer controlled videomicroscopy technique. The experimentally found universal exponential distribution of the velocity fluctuations is explained based on a simple, far-from equilibrium stochastic model. Moreover, in the early steps of in vitro neurogenesis from immortalized progenitor cells emergence of various self-organized structures has been observed, such as ordered migration or formation of dense aggregates. Some of these phenomena can be interpreted based on recent theoretical studies of collective motion of self-propelled, interacting particles.

9:30 AM K9.4 
THE INFLUENCE OF INTRACELLULAR TRANSPORT ON THE AXONAL AND DENDRITIC DEVELOPMENT OF NEURONAL FORM. H.G.E. Hentschel, Department of Physics, Emory University, Atlanta, GA; A. Fine, Department of Physiology and Biophysics, Faculty of Medicine, Dalhousie University, Halifax, Novia Scotia, CANADA.

We describe how both axonal differentiation and the dendritic branching during neuronal growth are dependent on intracellular transport. The interaction of diffusion and active transport of a morphogen whose concentration at the neurite tips influence the growth rate can lead to an instability involving winner-take-all growth, forming a single, quickly growing neurite (the axon) and inhibiting the growth of the other neurites. Ia addition, dendritic arbors have a growth and form consistent with the concept that shape is controlled by the local submembrane concentration of the calcium ion which is transported to its site of action via diffusion-controlled growth through biologically active membranes interacting with active pumping at the membrane surface.

9:45 AM K9.5 
A MODEL FOR DIFFUSION AND COMPETITION IN CANCER GROWTH AND METASTASIS. G.P. Pescarmona, Dipartimento di Genetica, Biologia e Chimica Medica, Universitá di Torino, Torino, ITALY; M. Scalerandi, P.P. Delsanto, INFM, Dipartimento di Fisica, Politecnico di Torino, ITALY; C.A. Condat, Department of Physics, University of Puerto Rico, Mayaguez, PR.

A master equation formalism is used to model the growth and metastasis of a tumor as a function of the diffusion and absorption of a nutrient. Healthy and cancerous (C-) cells compete to bind the nutrient, which is allowed to diffuse starting from a prescribed region. Two thresholds are defined for the quantity of nutrient bound by the C-cells. If this quantity falls below the lower threshold, the cell dies, while if it increases above the upper threshold, the cell divides according to a predefined stochastic mechanism. C-cells migrate when they record a low concentration of free nutrient in the local environment. The model is formulated in terms of a coupled system of equations for the cell populations and the free and bound nutrient. This system can be solved by using the Local Interaction Simulation Approach (LISA), a numerical procedure that permits an efficient and detailed solution and is easily adaptable to parallel processing. With suitable parameter variation, the model can describe multiple tumor configurations, ranging from the classical spheroid with a necrotic core favored by mathematicians to very anisotropic shapes with inhomogeneous concentrations of the various populations. This is important because the nature of the anisotropy may be crucial in determining whether and how the cancer metastasizes. The effects of stochasticity, due. e.g., to tissue inhomogeneity, and the presence of additional nutrients or inhibitors can be easily incorporated.

10:30 AM K9.6 
SPATIO-TEMPORAL EVOLUTION MODEL FOR COMPETITIVE LIGAND BINDING. C.A. Condat, Department of Physics, University of Puerto Rico, Mayaguez, PR; P.P. Delsanto, G.P. Perego, E. Ruffino, INFM, Dipartimento di Fisica, Politecnico di Torino, ITALY; D. Iordache, Department of Physics, Univ. Politehnica of Bucharest, Bucharest, ROMANIA.

Ligand binding is an essential step in many biological and biochemical processes, notably protein activation. In many cases, the binding rates are determined by competition - blockers competing with ligands or decoys competing with receptors. Often, the diffusivities of the various species also play a key role. The usual modelling of these processes, however, either involves well-mixed systems, for which spatial variations are neglected, or steady-state situations for which all time evolution has ceased. The model presented here can instead account for the spatial variations that originate as a consequence of inhomogeneous initial conditions and localized binding sites. It describes their evolution in time due to a combination of mechanisms of diffusion, competition and reaction. The model, which is formulated in terms of a coupled system of equations for the probability densities of the various populations, is easy to adapt to different specific conditions and is particularly well suited for numerical calculations, e.g., on a parallel computer. Since it yields concrete predictions, it can be readily used to predict the concentrations and distributions of the various species that are needed to obtain prescribed responses.

10:45 AM K9.7 
GENERATING AN INTRACELLULAR DIFFUSION MAP BY TWO-PHOTON FLUORESCENCE CORRELATION SPECTROSCOPY. Peter T. C. So, Keith Duggar, Dept. of Mechanical Engineering, MIT, Cambridge, MA; Keith M. Berland, TJ Watson Research Center, IBM, Yorktown Heights, NY; Erico Gratton, Dept. of Physics, UIUC, Urbana, IL.

Diffusion regulates intracelluar biochemical reaction rates by controlling transport processes. At present, the most commonly used technique to study intracellular diffusion is fluorescence photobleaching (FRAP). FRAP is a robust technique but suffers from the necessity of bleaching a large number of chromophores. It is possible that cellular transport may not be linear at high probe concentrations, or bleaching may caused cell damage via the generation of reactive oxygen intermediates. The recently developed two-photon fluorescence correlaton spectroscopy (FCS) technique allows intracellular diffusion to be measured at very low probe concentrations with a minimal photo energy deposition. The use of two-photon FCS to measure diffusion at selected locations in cells has been demonstrated. This presentaton describes an improvement on the basic technique by using a dual-color cross-correlation approach. This improved method enables the generation of a 3-D diffusion map throughout the intracellular volume.

11:00 AM K9.8 
MOVEMENT OF VESICLES IN EUCARIOTIC CELLS. ROLE OF INTRAVESICLE PROTONS AS A FUEL AND MODULATION OF THEIR CONCENTRATION BY DRUGS OR METABOLIC CHANGES. GianPiero Pescarmona, Emmanuella Morra, Elisabetta Aldieri, Amalia Bosia, Dipartimento di Genetica, Biologia e Chimica Medica, Universita' di Torino, Torino, ITALY.

Import (endocytosis) and export (secretion) of molecules from the cells is mediated by vesicles sliding along microtubules or actin filaments; a whole family of motor proteins has been described that are involved in membrane traffic. As the vescicles have been initially identified from the microscopic point of view they have received different names according to the cellular localization and function (lysosomes, Golgi complex, synaptic vesicles, etc). All these vescicles share a common feature: an internal pH of about 4, with a inner protons concentration 1000 fold higher than in the surrounding cytoplasm. As the protons gradient in mitochondria is able to drive ATP synthesis we can expect a similar role (energy supplier) for protons in all acidic vesicles. To experimentally test the vesicles mobility we have loaded them with a fluorescent dye and then measured the efflux of the dye over 6 hours. This efflux was reduced by all treatments lowering the actual concentration of protons in the vesicles, independently from the mechanism. Treatments included lowering of intracellular ATP or NADH, inhibitors of ATP- dependent proton translocase and/or the sodium/proton antiport, drugs that accumulate into lysosomes, buffering its acidity (chloroquine, doxorubicin). These results support the idea of a role of proton gradient as a fuel for protein motors in all cells tested. Moreover the similarity of the effects exerted on this system by many widely used drugs (cimetidine - antiacid, chloroquine -antimalarial and antiinflammatory, doxorubicin - chemotherapeutic, amphotericyn B -antibiotic, amiloride - diuretic) point to a lack of specificity of their action and can explain many of the side effect of these drugs.

11:15 AM K9.9 
ADHESION-INDUCED VESICLE PROPULSION. Isabelle Durand and Chaouqi Misbah, Laboratoire de Spectrometrue Physique, Universite Joseph-Fourier, Grenoble I, Saint-Martin d'Heres, FRANCE.

We study theoretically vesicle locomotion. We show how adhesion may lead to vesicle propulsion. The problem is fully solved numerically and an analytical solution is obtained in a perturbative scheme. The analytical result reproduces the numerical one. We provide an expression for the drift velocity as a function of relevant parameters. We discuss how a vesicle or a cell could establish itself this motion from physico-chemical concepts, while its environment could be initially homogeneous. We suggest experimental protocols.

SESSION K10: SINGLE MOLECULE MECHANICS AND DNA MANIPULATIONS 
Chairs: Bela M. Mulder and Helmut H. Strey 
Thursday Afternoon, December 4, 1997 
Grand Ballroom (S)

1:30 PM *K10.1 
SINGLE MOLECULE FORCE SPECTROSCOPY BY AFM - NANOMECHANICS MEETS MOLECULAR BIOLOGY. Hermann E. Gaub, Ludwig-Maximilians-Universitaet, Muenchen, GERMANY.

Recent developments in piconewton instrumentation allow the manipulation of single molecules and measurements of intermolecular as well as intramolecular forces. We took advantage of the high spatial resolution of the AFM and developed mechanical experiments with single macromolecules. An overview on this novel kind of spectroscopy will be given and applications in the field of polymer and life sciences will be highlighted: receptor ligand interactions were measured in single molecular pairs. Individual polymers and proteins that were anchored on a gold surface were picked up with the AFM tip and stretched, their viscolelasticity and yield strength was measured. Proteins were reversibly unfolded and the conformation forces were determined at the level of single secondary structure elements. A model was developed based on elastically coupled two-level systems that allows the description of basic features of the experimental results.

2:00 PM K10.2 
MICROSCOPIC ELASTICITY OF DNA FROM TORSIONALLY-CONSTRAINED STRETCHING. J. David Moroz and Philip Nelson, Physics and Astronomy, University of Pennsylvania, Philadelphia, PA.

In recent years it has become possible to pull single molecules with known forces and visualize the resulting conformational changes. This opens the door to extract microscopic elastic constants of DNA, which are useful in understanding its function, starting from macroscopic properties measurable with optical microscopy. Previous single-molecule work had only yielded the bend stiffness of the molecule (averaged over its two principal directions), the stretch constant, and a twist-stretch coupling. We explain the new phenomenon of torsionally-constrained entropic elasticity and show how to extract the twist stiffness and intrinsic rod anisotropy from the experiment of T. Strick et al., (Science 1996).

2:15 PM K10.3 
POLYMER PHYSICS WITH SINGLE MOLECULES OF DNA. Stephen Quake, Dept of Applied Physics, Caltech, Pasadena, CA; Hazen Babcock, Steven Chu, Dept of Physics, Stanford Univ, Stanford, CA.

Single molecules of DNA can be stained with fluorescent dye, visualized with an optical microscope, and manipulated with optical tweezers. This provides and excellent system for the study of polymer physics. In this talk, I will discuss some recent experiments with single molecules of DNA that test various aspects of the Zimm model of polymer dynamics, including the first direct measurement of a polymer's relaxation mode structure.

2:30 PM K10.4 
SEQUENCE SPECIFIC FORCE CURVES MEASURED BY MECHANICALLY OPENING THE DNA DOUBLE HELIX. Ulrich Bockelmann, Baptiste Essevaz-Roulet, Francois Heslot, LPMC, Ecole Normale Superieure, Paris, FRANCE.

Using techniques of molecular biology, we have designed a molecular construction which allows to attach the two complementary strands of one end of a single molecule of bacteriophage DNA separately to a glass microscope slide and a microscopic bead. A soft glass microneedle of calibrated stiffness (1.7 0.2 pN/m) is attached to the bead and imaged by an inverted optical microscope. Keeping the base of the microneedle fixed the glass slide is displaced laterally with constant velocity (20-200 nm/s). This leads to a progressive opening and unwinding of the double helix. We record and numerically analyze video sequences of the bead which allows to resolve subpiconewton variations in the force as a function of time. The force measured during the opening process shows a characteristic time variation which is determined by the base sequence in the vicinity of the advancing opening fork. Regions with higher content of G-C base pairs open at higher force than A-T rich regions. In the -phage DNA the G-C content varies between about 0.3 and 0.6 (average over 1000 base pairs) and we measure a force variation between 11.5 and 14 pN. Changing the direction of the displacement the double helix recombines and it is possible to perform several cycles of opening and closing on the same molecule. The sequence dependent force curves measured upon opening are reproducible to a high degree of detail. At a local scale the force signal shows clear sawtooth like structures. In a theoretical model, this is qualitatively explained as a molecular stick-slip motion which does not involve instabilities and is determined by the base sequence. Among exciting prospect of this new type of single molecule force measurement we see a quantitative investigation of the stacking and pairing interactions, studies of molecular stick-slip dynamics and also applications to the analysis of DNA or RNA sequences.

3:15 PM *K10.5 
PROSPECTS FOR SINGLE MOLECULE DNA SEQUENCING BY NANOPORES. Daniel Branton, John Kasianowicz, David Deamer, and Mark Akeson, Molecular and Cellular Biology, Harvard University, Cambridge, MA, Biotechnology Division, NIST, Gaithersburg, MD, and Dept. of Chemistry, UCSC, Santa Cruz, CA.

We are developing a single channel recording method that translates DNA bases into electronic signals directly and quickly and does so in a manner that is compatible with high levels of nano-fabrication. We believe these design criteria can be met by an instrument which draws single molecules of DNA through a small channel or pore that is integral to a sensitive detector. Thus far, we have been able to show that an electric field can drive single stranded RNA and DNA molecules through a . 2 nm diameter channel in a lipid bilayer membrane. Because the channel diameter was selected to accommodate only a single strand of RNA or DNA, each polymer is forced to traverse the membrane as an extended chain. During its traverse, the DNA partially blocks the channel, reducing ionic flow. As a result, the passage of each molecule is detected as a transient decrease of ionic current whose duration is proportional to polymer length and whose magnitude is dependent on the nature of the passing nucleotide. Channel blockades have been used to measure polynucleotide length and to distinguish between purines and pyrimidines. With further improvements, the method could in principle provide direct, high-speed detection of the sequence of bases in single molecules of DNA or RNA.

3:45 PM K10.6 
STRUCTURE AND INTERACTIONS IN SELF-ASSEMBLED DNA-CATIONIC LIPID COMPLEXES IN THE PRESENCE OF SIGLE AND MULTIVALENT COUNTERIONS. Ilya Koltover, Tim Salditt and Cyrus R. Safinya, Materials Department, Physics Department and Biochemistry and Molecular Biology Program, University of California at Santa Barbara.

When DNA is mixed with molecules of opposite charge, such as cationic lipids, multivalent cations, or polymers, it condenses in a variety of self assembled structures which may serve as a model for DNA condensation in-vivo. 123 We have established that linear DNA is condensed by the mixtures of cationic lipid (DOTAP) and a neutral lipid (DOPC), forming a multilayer liquid crystalline phase with DNA intercalated between the lipid bilayers in a periodic 2D smectic phase23. To elucidate the interactions driving the condensation and the complex structure we have conducted a detailed study of the phase diagram as a function of DOTAP/DNA (L+/D) and neutral/cationic lipid (v) ratios at different concentration M of monovalent counterions (salts). The condensation is driven by release of the lipid and DNA counterions and subsequent binding of lipids on the DNA strands. We present data on the dependence of the interaxial distance (dDNA) on the salt concentration M, which reveals the origin of the interactions between DNA strands. dDNA varies non-monotonically with M and depends non-linearly on the membrane charge density. Unexpectedly, we have found two different regimes of dDNA (M) variation at very low and high v ratios. We describe the differences in the effect of cations of different valence on the complex structure. Supported by NSF grant DMR-9624091, the Petroleum Research Fund (No.31352-AC7) and a Los Alamos CULAR grant No.STB/UC:95-146.

4:00 PM K10.7 
FORCES BETWEEN CATIONIC TRANSFECTION RELATED SURFACTANTS. Samuel Campbell, Chad Park, Dan Lasic, Jacob Israelachvili, Department of Chemical Engineering, University of California, Santa Barbara, CA.

Experiments have been performed on the liposome-forming surfactant, DODAB (cationic) with and without the presence of DOPE (zwitterionic). The combinations are used as a DNA transfection agent. The surface forces apparatus (SFA) was utilized to measure the forces between Langmuit-Blodgett bilayers exposing DODAB in the outer monolayers under varying salt concentrations with and without DNA (single-stranded). These represent the first measurement of forces between surfactant/DNA assemblies. The forces without DNA are found to be repulsive due to electrostatic interactions at all measured salt concentrations. In contrast, the forces between bilayers in the presence of DNA changed from a purely repulsive interaction due to DNA absorption at I mM and 10 mM salt to an attractive and adhesive bridging interaction at 150 mM salt. These results are consistent with recently reported X-ray measurements. Understanding the mechanism of DNA transfection, particularly, the adhesion of the DNA/surfactant assemblies to cells prior to the DNA exchange is the prime motivation for this study.

4:15 PM K10.8 
QUASI 2-D SMECTIC STATES OF DNA MOLECULES INTERCALATED BETWEEN LIPID MEMBRANES IN MULTI-LAMELLAR PHASES. L. Golubovic, D. Moldovan and M.Golubovic, West Virginia University, Morgantown, WV.

DNA-cationic lipid complexes are promising carriers of DNA in gene therapy. These complexes form lamellar lipid membrane phases with DNA intercalated between membranes.[l] DNA strands form striking first example of strongly fluctuating quasi 2-d smectic phases only weakly interacting across lipid membranes.[l] Experimentally, these quasi 2-d smectic phases exhibit only a short-range positional order due to divergent fluctuations of DNA in-plane displacements similar to undulations of true 2-d smectics [2] (as if the DNA couplings across the membranes would be zero). Here, we theoretically discuss effects of the couplings between adjacent smectic planes as their presence has been evidenced in experiments.[l] Weak couplings do not restore DNA positional order as DNA thermal undulations,  z^2 br z with_ behavior, smectic 2-d ordinary observed[l] experimentally the them, to Due scales. length long at relevant be out turn couplings interplane However, planes. zandxof sizes of power-laws as grow still z, z%%>2> 
z^1/2x, occurs only at short enough scales. At longer scales we predict a novel fluctuation behavior, with 2> 
(ln z)^2(ln x)^2.

4:30 PM K10.9 
SEMI-FLEXIBLE POLYMERS GRAFTED TO SURFACES: TOROIDAL MICELLES AND COLLAPSE INSTABILITIES. Joanne Bright, David R.M. Williams, Australian National University, Research School of Physical Sciences, Canberra , AUSTRALIA.

There has over the past decade been a great deal of work on the grafting of fully-flexible polymer chains to surfaces. However, many polymers of biological origin are semi-flexible and require significant energy to form a bend. Here we examine how such polymers behave when end-grafted to a surface in the presence of a poor solvent. We show that in the low-density regime the polymers will form toroidal surface micelles, as is seen experimentally in DNA. The toroids represent a balance between the tendency of the surface area to be minimised and the energetic cost of bending the chains. It is also possible for tower-like micelles to form. In the high-density limit there is a small regime where the surface undergoes an undulational instability. However the most common form of instability is uniform collapse of the layer by all the chains bending in one direction.

4:45 PM K10.10 
DNA TOROIDAL CONDENSATION: A THEORETICAL MODEL. Stella Y. Park, William M. Gelbart, University of California, Los Angeles, Dept of Chemistry and Biochemistry, Los Angeles, CA.

DNA molecules in dilute solutions are known to condense reversibly into compact toroids, whose size is essentially independent of: condensing agent (multivalent cations, neutral or charged polymer, or alcohol); DNA base-pair sequence and molecular weight (over a range from 400 to 50,000 base-pairs); and background salt concentration. We propose that this size invariance is due to the presence of topologically unavoidable packing defects associated with the circumferential winding of DNA strands into toroids. In particular, we show that a simple model for these defects and their energies can account for a unique size of the torus.