TISSUE BIOPHYSICS AND BIOMIMETICS
, Head, Section on Tissue Biophysics and Biomimetics
Ferenc Horkay, PhD, Staff Scientist
Carlo Pierpaoli, MD, PhD, Staff Scientist
Lin-Ching Chang, PhD, Postdoctoral Fellow
Cheng Guan Koay, PhD, Postdoctoral Fellow
David Lin, PhD, Postdoctoral Fellow
Joelle Sarlls, PhD, Postdoctoral Fellow
Michal Komlosh, PhD, Visiting Associate
Uri Nevo, PhD, Visiting Fellow
Evren Özarslan, PhD, Visiting Fellow
Iren Horkayne-Szakaly, MD, Volunteer
Candida Silva, PhD, Volunteer
Ichiji Tasaki, MD, PhD, Volunteer
Lindsay Walker, MS, Volunteer
We try to understand fundamental relationships between function and structure in soft tissues, in “engineered” tissue constructs, and in tissue analogues, focusing on how microstructure, hierarchical organization, composition, and material properties of tissue affect biological function and dysfunction. Our biophysical perspective involves physical measurements in tandem with analytical and computational models in order to interrogate biological and physical model systems at different time and length scales. Primarily, we use water to probe both equilibrium and dynamic interactions among tissue constituents over a wide range of these scales. To determine the equilibrium osmo-mechanical properties of well-defined model systems, we systematically vary water content or ionic composition. To probe tissue structure and dynamics, we employ small-angle X-ray and neutron scattering, static and dynamic light scattering, and nuclear magnetic resonance (NMR) relaxometry. We use mathematical models to understand how changes in tissue microstructure and physical properties affect essential transport processes (e.g., of mass, charge, momentum). The most direct noninvasive means to characterize these transport processes in vivo is with magnetic resonance imaging (MRI) methods, which can characterize normal tissue microstructure and follow its changes in development, degeneration, and aging. Another goal is to translate our new quantitative methodologies from “bench to bedside.”
Virtual in vivo tissue biopsy
We are continuing to develop novel magnetic resonance (MR) methods that allow us to probe tissue microstructure and diagnose neurological and developmental disorders in vivo. Diffusion Tensor MRI (DT-MRI) is the most mature such technology that we have invented and developed. With it, we measure a diffusion tensor of water, D, on a pixel-by-pixel basis within tissue. Information derived from D includes the local fiber-tract orientation, the mean-squared distance that water molecules diffuse in any given direction, the orientationally averaged mean diffusivity, and other intrinsic scalar invariant quantities that are independent of the laboratory coordinate system. These scalar parameters are like quantitative histological “stains” that are “developed” without exogenous contrast agents or dyes. The bulk or orientationally averaged diffusivity has been the most successful imaging parameter proposed to date for identifying ischemic tissue regions in the brain during acute stroke. Furthermore, measures of diffusion anisotropy are useful in identifying white matter degeneration (e.g., Wallerian degeneration) associated with chronic stroke. DT-MRI also provides new information about early developmental changes in cortical gray and white matter that cannot be detected with other imaging methods.
We have also pioneered the use of color to encode nerve fiber orientation, allowing us to identify the main association, projection, and commissural white matter pathways in the brain and even to differentiate white matter pathways with similar structure and composition but distinct spatial orientations. To assess anatomical connectivity between different functional brain regions, we developed DT-MRI fiber tractography to track nerve fiber tract trajectories by continuously following the direction along which the apparent diffusivity is a maximum. While the reliability of DT-MRI tractography is high in coherently organized primary white matter pathways, it is possible to generate artifactual tracts when following white matter pathways with a more complex underlying fiber topology (e.g., that cross, merge, or branch). To assess the reliability of individual computed tracts, we have used point-wise assessment of streamline tractography attributes (PASTA). We are using other MR data and more detailed models of water diffusion in tissue to compensate for artifacts.
Recently, we proposed several advanced in vivo MR methods to measure fine microstructural features of nerve fascicles that previously could be measured only by using laborious histological methods. One approach is a combined experimental and modeling framework (composite hindered and restricted model of diffusion or CHARMED). By enabling us to describe water diffusion in white matter, CHARMED permits us to distinguish between water diffusing within the intra-axonal spaces and extra-axonal compartments, thereby improving the angular resolution in tract tracing and resolution of fibers that cross. More recently, we extended CHARMED to measure the axon diameter distribution within large fascicles, a method we call Axcaliber. We eventually hope to use the resultant information to assess pathological changes in white matter in disease and to segment white matter regions in the brain and spinal cord according to the local diameter distribution of axons.
Our group expended significant effort to develop advanced MR methods to probe correlations between displacements of water molecules within brain tissue. A primary target is gray matter, which appears featureless in DT-MR images but whose microarchitecture is rich and varied throughout the brain. We have been developing specialized MR sequences to probe different gray matter compartments at microscopic resolution, as well as detailed mathematical models describing water displacements therein, in order to infer or estimate microstructural and morphological features.
In general, clinical and biological applications of DT-MRI (and of other novel displacement MRI methods) require the development of new mathematical, statistical, and image-sciences concepts and tools in order to permit analysis of novel multidimensional data sets. We have developed algorithms for continuous, smooth approximation to the discrete, noisy, measured diffusion tensor field data, reducing noise and enabling us to follow fibers more reliably. We recently derived a new Gaussian distribution for tensor-valued random variables, which we used in designing “optimal” DT-MRI experiments. In addition, we have developed nonparametric empirical (e.g., Bootstrap) methods for determining features of the statistical distribution of experimental DT-MRI data and, most recently, for studying the inherent variability and reliability of white matter fiber tract trajectories. These parametric and nonparametric statistical methods allow us to apply powerful statistical hypothesis tests to a wide range of important biological and clinical questions that previously could be examined only with ad hoc methods. More recently, we have been developing a variety of empirical and perturbation methods to describe the effect of noise on DT-MRI data, permitting assessment of the variability of important tensor-derived quantities that are now widely used in clinical practice.
Overall, we continue to use MR water diffusion data to perform “virtual in vivo tissue biopsy” of the brain; we obtain detailed microstructural or anatomic features not available with other methods. The data permit us to characterize normal and abnormal brain development, parcellate different cortical regions, and assess changes associated with disease, degeneration, and aging.
Assaf Y, Basser PJ. Composite hindered and restricted model of diffusion (CHARMED) MR imaging of the human brain. Neuroimage 2005;27:48-58.
Koay CG, Basser PJ. Analytically exact correction scheme for signal extraction from noisy magnitude MR signals. J Mag Res 2006;179:317-22.
Koay CG, Carew JD, Alexander AL, Basser PJ, Meyerand ME. Investigation of anomalous estimates of tensor-derived quantities in diffusion tensor imaging. Magn Reson Med 2006;55:930-6.
Özarslan E, Basser PJ, Shepherd TM, Thelwall PE, Vemuri BC, Blackband SJ. Observation of anomalous diffusion in excised tissue by characterizing the diffusion-time dependence of the MR signal. J Magn Reson 2006;183:315-23.
Pajevic S, Aldroubi A, Basser PJ. Continuous tensor field approximation of diffusion tensor MRI data. In: Weickert J, Hagen H, eds. Visualization and Processing of Tensor Fields. Springer, 2006;299-314.
MRI study of normal brain development
Four NIH Institutes (NICHD, NIMH, NINDS, and NIDA) have jointly sponsored a multicenter study to advance our understanding about how the brain develops in typical, healthy children and adolescents. The study has enrolled approximately 500 children, ranging from infancy to young adulthood, who have been seen at different time points over a six-year period (2001–2007) at several clinical centers around the United States. MRIs of their brains have been acquired, and the imaging findings will be related to the results of standardized neuropsychological tests.
Our role in the project is to serve as a Diffusion Tensor Data Processing Center (DPC). We process and analyze all diffusion tensor MRI (DTMRI or DTI) data acquired by the various Centers during the course of the study. Previously, we developed and implemented the DTI data processing pipeline. Specifically, we developed procedures for sorting, displaying, and co-registering diffusion-weighted images from which the DTI data are computed. The image registration procedure accomplishes the tasks of removing the effects of subject motion and eddy current distortion and of aligning the images to a given template with only one interpolation step, thus ensuring minimal loss of data quality. We also proposed a new strategy for robust estimation of the diffusion tensor and quantities derived from it. We addressed the task of registering DTI data to other structural MRI data contained in the database by using rigid body and linear (affine) transformations. More recently, we addressed the issue of correcting residual image distortion originating from the Echo-Planar Image (EPI) acquisition used for DTI.
In 2005, the project was awarded Neuroscience Blueprint funds to expand the DTI portion of the investigation. The original DTI scan was limited to a 10-minute protocol performed on those subjects able to tolerate additional time in the scanner following the collection of the core structural MRI data. The expanded study will increase the quality and quantity of data collected through longer or additional scanning sessions for older subjects and through the recruitment of new subjects and scanning sessions for the younger cohort. Our group has designed the expanded DTI protocol and coordinated its implementation at each acquisition site. Data collection was completed; we expect that all raw diffusion data and structural MRI targets will soon be available for systematic processing. We plan to produce processed diffusion tensor data for dissemination to the scientific community about a year after receiving all the raw data.
Almli CR, Rivkin MJ, McKinstry RC, Brain Development Cooperative Group (Pierpaoli C, member). The NIH MRI study of normal brain development (Objective-2): newborns, infants, toddlers, and preschoolers. Neuroimage 2007;35:308-25.
Brain Development Cooperative Group (Pierpaoli C, member), Evans AC. The NIH MRI study of normal brain development. Neuroimage 2006;30:184-202.
Chang LC, Koay CG, Pierpaoli C, Basser PJ. Variance of estimated DTI-derived parameters via first-order perturbation methods. Magn Reson Med 2007;57:141-9.
Jones DK, Catani M, Pierpaoli C, Reeves SJ, Shergill SS, O’Sullivan M, Golesworthy P, McGuire P, Horsfield MA, Simmons A, Williams SC, Howard RJ. Age effects on diffusion tensor magnetic resonance imaging tractography measures of frontal cortex connections in schizophrenia. Hum Brain Mapp 2006;27:230-8.
Koay CG, Chang LC, Carew JD, Pierpaoli C, Basser PJ. A unifying theoretical and algorithmic framework for least squares methods of estimation in diffusion tensor imaging. J Magn Reson 2006;182:115-25.
Biopolymer physics: water-ion-biopolymer interactions
To help us understand the nature of physical/chemical interactions taking place in biomolecules and biomolecular assemblies in living systems, we developed an experimental approach for studying biomolecules’ and assemblies’ structure (morphology) and thermodynamic properties as a function of the length scale (i.e., spatial resolution). The methodology combines macroscopic osmotic swelling pressure measurements and various high-resolution scattering experiments. Macroscopic swelling pressure measurements probe in the coarse length–scale regime, providing information about overall thermodynamic response. Small-angle neutron scattering (SANS), static light scattering (SLS), and dynamic light scattering (DLS) collectively provide information about the sizes of structural elements and their respective contribution to macroscopic osmotic properties. Combining these measurements allows us (1) to separate the scattering intensity caused by thermodynamic concentration fluctuations from large static superstructures (e.g., aggregates) and (2) to determine the length scales of fluctuations that give rise to macroscopic thermodynamic properties.
Moreover, in solutions of large charged polymers, counterions form a cloud surrounding the polymer chain. With increasing salt concentration, a fraction of these ions adsorb onto the macroion. It is known that the distribution of monovalent ions can be approximated by the Poisson-Boltzmann theory. This theory, however, fails to describe the counterion atmosphere in the presence of multivalent counterions whose effect on the chain conformation also remains poorly understood. Knowledge of the counterion distribution is therefore essential to understanding counterions’ biological function. Recently, we developed a method based on the measurement of the X-ray scattering signal in the vicinity of the absorption edge of the counterions that allows us to determine the distribution of ions around charged biopolymer molecules.
Specifically, we have applied the above multiscale approach to the effects of multivalent cations, particularly calcium, on the structure and morphology of various biomolecules. Divalent cations are ubiquitous in the biological milieu, yet existing theories do not adequately explain their effect on and interactions with charged polymers or biomolecules. Moreover, experiments to study the interactions are difficult to perform, particularly in solution, because multivalent cations above a low concentration threshold generally cause phase separation or precipitation of charged molecules. Given that macroscopic phase separation does not occur in crosslinked gels, we overcame the limitation associated with concentration levels by crosslinking our biopolymers, thereby extending the range of ion concentrations over which the system remains stable and may be studied. We applied the new method in pilot studies of DNA and of crosslinked gels of the model synthetic polymer polyacrylic acid. Concentrated DNA solutions and novel DNA gels have never before been investigated with SANS in conjunction with osmotic measurements. Our method also provides a unique framework for analyzing the osmotic and scattering behavior of other biomolecular systems.
We also investigated crosslinked gels of the biopolymer hyaluronic acid to determine the size of the structural elements that contribute to the osmotic concentration fluctuations. We combined small-angle X-ray scattering (SAXS) and SANS to estimate the osmotic modulus of hyaluronic acid solutions in the presence of monovalent and divalent counter-ions. Using DLS to determine the osmotic modulus independently from the dynamic response, we also investigated the diffusion processes in biopolymer solutions.
Horkay F, Basser PJ. Effect of ions on the thermodynamic properties of biopolymer gels. Polym Mater Sci Eng 2007;96:950-1.
Horkay F, Han MH, Han IS, Kim J, Bang IS, Magda JJ. Separation of the effects of pH and polymer concentration on the swelling pressure and elastic modulus of a pH-responsive hydrogel. Polymer 2006;47:7335-9.
Horkay F, Hecht A, Geissler E. Similarities between polyelectrolyte gels and biopolymer solutions. J Polymer Sci B Polymer Phys 2006;44:3679-86.
Horkay F, McKenna GB. Polymer networks and gels. In: Mark JE, ed. Physical Properties of Polymers Handbook. Springer, 2007;497-523.
Van Thienen TG, Horkay F, Braeckmans K, Stubbe BG, Demeester J, De Smedt SC. Influence of free chains on the swelling pressure of PEG-HEMA and dex-HEMA hydrogels. Int J Pharm 2007;337:31-9.
Functional properties of extracellular matrix
The swelling behavior of cartilage is exquisitely sensitive to biochemical and microstructural changes occurring in development, disease, degeneration, and aging. Given that it is primarily the osmotic properties of cartilage’s constituents that allow cartilage to resist applied loads and express or imbibe fluid, respectively, knowledge of the mechanisms affecting cartilage hydration (water uptake) is also essential to understanding and predicting cartilage’s load-bearing and lubricating ability.
Conversely, we previously showed that controlled hydration or swelling of cartilage can provide a means of determining functional properties of cartilage’s extracellular matrix (ECM), thus permitting independent measurement of important physical/chemical properties of the collagen network and proteoglycans (PG) within the ECM. This approach involves (1) modeling the ECM as a composite material consisting of two distinct phases: a collagen network and a proteoglycan solution trapped within it; (2) applying various known levels of equilibrium osmotic stress; and (3) using physical-chemical principles and additional experiments to determine a “pressure-volume” relationship independently for both the PG and collagen phases. In pilot studies, we used this approach to determine pressure-volume curves for the collagen network and the PG phases in native and trypsin-treated normal human cartilage specimens as well as in cartilage specimens from osteoarthritic (OA) joints. In both normal and trypsin-treated specimens, collagen network stiffness appeared unchanged, whereas collagen network stiffness was lower in the OA specimen. Our findings highlight the role of the collagen network in limiting normal cartilage hydration and ensuring a high PG concentration in the matrix, both of which are essential for effective load bearing in cartilage and lubrication but are lost in OA. The data also suggest that the loss of collagen network stiffness, not the loss or modification of PGs, may be the incipient event leading to the subsequent disintegration of cartilage observed in OA.
One shortcoming of our approach, however, was that it takes many days to analyze a single cartilage specimen, making the technique unsuited to routine pathological analysis or tissue engineering applications. Furthermore, a significant amount of tissue is required to perform the osmotic titration experiments. Therefore, we recently invented and developed a new tissue micro-osmometer to perform such experiments practically and rapidly. The instrument can measure minute amounts of water absorbed by small tissue samples (less than 1 µg) as a function of the equilibrium activity (pressure) of the surrounding water vapor. A quartz crystal detects the water uptake of a specimen attached to the crystal’s surface. The high sensitivity of the crystal’s resonance frequency to small changes in the amount of adsorbed water allows us to measure precisely the mass of water taken up by the tissue specimen. Varying the equilibrium vapor pressure surrounding the specimen induces controlled changes in the osmotic pressure of the tissue layer.
To illustrate the applicability of the new apparatus, we measured the swelling pressure of a tissue-engineered cartilage specimen. In tandem, we used atomic force microscopy (AFM) to determine the local mechanical properties. The concentration of the main biopolymer components is determined by biochemical analysis. We also studied the contribution from individual components to the tissue’s osmotic and mechanical properties. We are now focusing on aggrecan, which is the most abundant cartilage proteoglycan. We study the static properties of aggrecan and aggrecan/hyaluronic acid assemblies by osmotic pressure measurements, SLS, SANS, and SAXS while probing the dynamics by DLS and neutron spin-echo measurements.
The micro-osmometer will eventually permit us to obtain a profile of the osmotic compressibility or stiffness of several cartilage specimens simultaneously as a function of depth from the articular surface to the bone interface. It will also allow us to quantify the contributions of individual components of ECM (e.g., aggrecan, hyaluronic acid, collagen) to total osmotic pressure. Moreover, it should allow us to assess the osmotic compatibility and mechanical integrity of developing tissues and tissue-engineered cartilage (or ECM) for improved integration following implantation.
We recently developed an AFM technique for mapping the local elastic properties of tissues. We have successfully addressed many of the issues that previously hindered use of AFM in high-throughput probing of inhomogeneous samples, particularly biological tissues. The technique uses the precise scanning capabilities of a commercial AFM to generate large volumes of compliance data and automatically extracts the relevant elastic properties from the data. In conjunction with results obtained from micro-osmometry and biochemical analysis, AFM will allow us to map spatial variations in the load-bearing capacity of cartilage specimens.
Horkay F, Hecht AM, Rochas C, Basser PJ, Geissler E. Anomalous small angle x-ray scattering determination of ion distribution around a polyelectrolyte biopolymer in salt solution. J Chem Phys 2006;125:234904.
Langrana N, Horkay F, Yurke B, eds. Biomimetic Polymers and Gels. MRS Symposium Series, Volume 897E. Materials Research Society, 2005.
Lin DC, Dimitriadis EK, Horkay F. Advances in the mechanical characterization of soft materials by nanoindentation. In: Pandalai SG. Recent Research Developments in Biophysics, vol. 5, part II. Transworld Research Network (Kerala), 2006;333-70.
Lin DC, Dimitriadis EK, Horkay F. Robust strategies for automated AFM force curve analysis–I. Non-adhesive indentation of soft, inhomogeneous materials. J Biomech Eng 2007;129:430-40.
Michelman-Ribeiro A, Horkay F, Nossal R, Boukari H. Probe diffusion in aqueous poly(vinyl alcohol) solutions studied by fluorescence correlation spectroscopy. Biomacromolecules 2007;8:1595-600.
COLLABORATORS
Akram Aldroubi, PhD, Vanderbilt University, Nashville, TN
Yaniv Assaf, PhD, Tel Aviv University, Tel Aviv, Israel
Alan Barnett, PhD, Clinical Brain Disorders Branch, NIMH, Bethesda, MD
Yoram Cohen, PhD, Tel Aviv University, Tel Aviv, Israel
Emilios Dimitriadis, PhD, Division of Bioengineering and Physical Science, ORS, OD, NIH, Bethesda, MD
Raisa Freidlin, MS, Computational Bioscience and Engineering Laboratory, CIT, NIH, Bethesda, MD
Erik Geissler, PhD, Laboratoire de Spectrométrie Physique, CNRS, Université Joseph Fourier de Grenoble, Grenoble, France
Anne-Marie Hecht, PhD, Laboratoire de Spectrométrie Physique, CNRS, Université Joseph Fourier de Grenoble, Grenoble, France
Wolfram Jarisch, PhD, LifeStar LLC, Potomac, MD
Derek Jones, PhD, Cardiff University, Cardiff, Wales, UK
Stefano Marenco, MD, Clinical Brain Disorders Branch, NIMH, Bethesda, MD
Pedro Miranda, PhD, Universidade de Lisboa, Lisbon, Portugal
Sinisa Pajevic, PhD, Mathematical and Statistical Computing Laboratory, CIT, NIH, Bethesda, MD
Valery Pikalov, PhD, Institute of Theoretical and Applied Mechanics of the Russian Academy of Sciences, Novosibirsk, Russia
Pabitra Sen, PhD, Schlumberger-Doll Research, Cambridge, MA
The Brain Development Cooperative Group
For further information, contact pjbasser@helix.nih.gov.
