High rate shear strain of three-dimensional neural cell cultures: a new in vitro traumatic brain injury model

https://doi.org/10.1016/j.jbiomech.2004.05.032Get rights and content

Abstract

The fidelity of cell culture simulations of traumatic brain injury (TBI) that yield tolerance and mechanistic information relies on both the cellular models and mechanical insult parameters. We have designed and characterized an electro-mechanical cell shearing device in order to produce a controlled high strain rate injury (up to 0.50 strain, 30 s−1 strain rate) that deforms three-dimensional (3-D) neural cultures (neurons or astrocytes in an extracellular matrix scaffold). Theoretical analysis revealed that these parameters generate a heterogeneous 3-D strain field throughout the cultures that is dependent on initial cell orientation within the matrix, resulting in various combinations of normal and shear strain. The ability to create a linear shear strain field over a range of input parameters was verified by tracking fluorescent microbeads in an acellular matrix during maximal displacement for a range of strains and strain rates. In addition, cell death was demonstrated in rat cortical astrocytes and neurons in response to high rate, high magnitude shear strain. Furthermore, cell response within the 3-D neuronal cultures depended on orientation, with higher predicted shear strain correlating with an increased loss of neurites, indicating that culture configuration may be an important factor in the mechanical, and hence cellular, response to traumatic insults. Collectively, these results suggest that differential responses exist within a 3-D culture subjected to mechanical insult, perhaps mimicking the in vivo environment, and that this new model can be used to investigate the complex cellular mechanisms associated with TBI.

Introduction

Traumatic brain injury (TBI) is a severe health and socioeconomic problem, for which there are few effective clinical treatments (Roberts et al., 1998). TBI results from mechanical loading to the head and therefore models that seek to reveal injury mechanisms should accurately simulate the related biomechanics. Traumatic loading to the head can involve several components, including contact and/or inertial loading (Gennarelli, 1993). The severity of neurological disability depends on the initial insult and the ensuing cellular cascades, which may be complex and persistent.

Although in vivo studies model the systemic and behavioral deficits of TBI, in vitro approaches provide a powerful framework for investigating isolated mechanisms. Cellular models range from transection (Gross et al., 1983; Lucas et al., 1985; Mukhin et al., 1996; Regan and Choi, 1994; Tecoma et al., 1989) and compression models (Balentine et al., 1988; Murphy and Horrocks, 1993; Shepard et al., 1991; Wallis and Panizzon, 1995) (contact loading) to more complex stretching and acceleration devices (Cargill and Thibault, 1996; Ellis et al., 1995; Geddes and Cargill, 2001; LaPlaca and Thibault, 1997) (inertial loading). Inertial loading has been correlated with severe, diffuse brain injury that is attributed to stretching neural components (Gennarelli, 1993; Margulies and Thibault, 1992; Ommaya et al., 1994). The spatial and temporal patterns of strain associated with inertial loading are the basis for several in vivo and in vitro injury models that apply large strains (see Gennarelli, 1994; Morrison et al (1998a), Morrison et al (1998b) for reviews). In addition to the strain magnitude, the tensorial nature of strain is a determining factor in tissue outcome. In fact, the primary mode of failure in neural tissue is shear (Holburn, 1943; Shuck and Advani, 1972). Models that employ rapid mechanical deformation often exhibit a strain- and/or strain rate-dependent injury, thus substantiating the clinical observation that angular acceleration and injury severity are causally related.

Planar (2-D) cultures differ markedly from the three-dimensional (3-D) cytoarchitecture of the brain. Cell behaviors of 2-D versus 3-D cell configurations have shown that 3-D cultures more closely resemble those of native tissue (e.g., osteoblastic cells (Granet et al., 1998), hepatocytes (Takeshita et al., 1998), breast cells (Wang et al., 1998), and neural cells (Fawcett et al (1989), Fawcett et al (1995))), suggesting that 3-D models may yield more accurate secondary responses. Cells cultured in 2-D have different cell–cell and cell–matrix interactions than 3-D cultures (Cukierman et al (2001), Cukierman et al (2002); Gumbiner and Yamada, 1995; Schmeichel and Bissell, 2003; Yamada et al., 2003), potentially impacting mechanotransduction associated with traumatic insults. In this context, brain slices (ranging in thickness from 200 to 500 μm) have been utilized to determine molecular responses to injury (Morrison et al (1998a), Morrison et al (1998b); Sieg et al., 1999; Wallis and Panizzon, 1995). Although slice models are invaluable to in vitro investigations, the ability to control cell types, ratio of cell types, and extracellular components—factors that may be important to the injury response—is limited.

A device that incorporates the features of 3-D cell cultures under conditions of simple shear strain would be a significant addition to the repertoire of cell injury devices in the study of TBI. To this end, we have developed a 3-D cell shearing device (CSD) that delivers a prescribed shear strain to 3-D cell cultures. The strain is controlled by a closed loop system, ensuring precise and repeatable deformation. The objectives of the current study are (1) to analytically describe the 3-D strain field for a range of shear displacement angles, (2) to measure the strain field throughout a 3-D gel configuration, and (3) to test the ability of the system to induce cell injury in 3-D neural cultures (neonatal rat cortical astrocytes and embryonic rat cortical neurons).

Section snippets

Three-dimensional cell shearing device (3-D CSD)

The primary design goal for the 3-D CSD was to have control over the magnitude and rate of displacement. The 3-D CSD has two major components: the cell chamber and the actuator/control system (Fig. 1a). The cell chamber is designed to contain enough cells for molecular analyses in a 3-D configuration. The actuator/control system develops shear strains up to 0.50 (shear angle up to 45°) at rates from 1 to 30 s−1 in order to simulate the spatial and temporal strain patterns associated with

Kinematic analysis of 3-D strain field

It was hypothesized that the orientation of the cell within the 3-D matrix contributes to the strain transferred to the individual cells from bulk deformation. It was assumed that the overall shear strain field was homogeneous, isotropic, and the culture was incompressible. To describe the motion of points within the deformable matrix, a fixed reference frame in 3-D space was established where the Cartesian coordinates of a point were denoted by x=(x1,x2,x3). The positions of any two adjacent

Discussion

We have designed a device that is capable of mechanically deforming 3-D neural cell cultures and demonstrated its use as a quantifiable, reproducible model of in vitro traumatic injury. A shear deformation was achieved through the parallel motion of the top plate of a cell chamber with respect to the bottom surface, thus uniformly deforming the 3-D cell culture. A linear shear strain field for several combinations of strain and strain rate was determined using microbead tracking, illustrating

Acknowledgements

Funding for this study was partially provided by the NSF (CAREER Award to ML, BES-0093830) and the NIH (EB001014-2). The authors also acknowledge Maggie Wolfson (GT, Department of Biomedical Engineering) for technical assistance in image analysis.

References (41)

  • R.A. Wallis et al.

    Felbamate neuroprotection against CA1 traumatic neuronal injury

    European Journal of Pharmacology

    (1995)
  • J.D. Balentine et al.

    In vitro spinal cord trauma

    Laboratory Investigation

    (1988)
  • R.S. Cargill et al.

    Acute alterations in [Ca2+]i in NG108-15 cells subjected to high strain rate deformation and chemical hypoxiaan in vitro model for neural trauma

    Journal of Neurotrauma

    (1996)
  • E. Cukierman et al.

    Taking cell-matrix adhesions to the third dimension

    Science

    (2001)
  • E. Cukierman et al.

    Cell interactions with three-dimensional matrices

    Current Opinion in Cell Biology

    (2002)
  • B.R. Donnelly et al.

    Shear properties of human brain tissue

    Journal of Biomechanical Engineering

    (1997)
  • E.F. Ellis et al.

    A new model for rapid stretch-induced injury of cells in culturecharacterization of the model using astrocytes

    Journal of Neurotrauma

    (1995)
  • J.W. Fawcett et al.

    Dopaminergic neuronal survival and the effects of bFGF in explant, three dimensional and monolayer cultures of embryonic rat ventral mesencephalon

    Experimental Brain Research

    (1995)
  • L.A. Flanagan et al.

    Neurite branching on deformable substrates

    Neuroreport

    (2002)
  • D.M. Geddes et al.

    An in vitro model of neural traumadevice characterization and calcium response to mechanical stretch

    Journal of Biomechanical Engineering

    (2001)
  • Cited by (173)

    • A computational investigation of strain concentration in the brain in response to a rapid temperature rise

      2021, Journal of the Mechanical Behavior of Biomedical Materials
      Citation Excerpt :

      These prior studies were focused on gaining insight on the frequency response of pressure waves, and did not examine brain tissue strains following larger, yet survivable, applied temperatures. Since strain has previously been empirically associated with injury thresholds in brain tissue (Bain and Meaney, 2000; LaPlaca et al., 2005; Cullen et al., 2007; Wright and Ramesh, 2012), here we use a finite element model to investigate the displacements and strains that result from exposure to pulsed microwaves. Personal accounts from investigators involved in studies of MAE on human subjects suggest that a small percentage of subjects might have experienced symptoms similar to mild closed head injury (Frey, 1988).

    View all citing articles on Scopus
    1

    Present address: Exponent Failure Analysis Associates, 3401 Market Street, Suite 300, Philadelphia, PA 19104, USA

    View full text