Multiscale Mechanobiology of Brain Injury: Axonal Strain Redistribution

Main Text

Those staying abreast of recent literature will recall that in 1803, Simmons (1) noted that “the term concussion conveys not a precise idea of that derangement which is produced in the organization of the brain by external violence.” The lack of precision persists even to this day, with the cellular and molecular mechanisms of mild traumatic brain injury (mTBI), as well as the thresholds of “external violence” that cause mTBI, still unresolved topics of debate. A recent insightful study published in Biophysical Journal (2) suggests that anatomical “derangement,” in the form of interfaces between gray and white matter, is an important determinant of how “external violence,” in the form of head acceleration, contributes to the onset of mTBI.

TBI is an important public health problem arising from a variety of causes, including sports-related impacts, motor vehicle accidents, falls, and assault. Of the estimated annual TBI occurrences in the United States alone, 75% of them can be classified as concussion or mTBI (3). Although 20–30% of mTBI patients suffer from persistent neurocognitive dysfunction, a large fraction of them remains unrecognized and unreported, in part because the associated symptoms are difficult to detect. This highlights the importance of 1) improved understanding of how external energy is transmitted from the skull to brain through meninges at the interface (4) across length scales and at the axon level (2) and 2) the development of reliable criteria for mTBI predictions in the real world (5).

One of the most common pathological features of mTBI is diffuse axonal injury (DAI), which causes delayed responsiveness and memory dysfunction. Although axons rarely disconnect at the time of injury, it is thought that the very long and thin structure of axons renders them susceptible to damage from stretches caused by the rapid brain deformation during head rotation (6). In fact, recent in vitro studies show that high and/or rapid tensile forces on micropatterned neuronal cells can break the core axonal microtubules and trigger axonal injury (7). Periodic swellings or bulbs were observed in these axonal injuries that are similar to that in human DAI. These experimental studies indicate that energy from the head impact is transmitted through the length scales of the brain to individual axons to cause damage. Nevertheless, the hierarchical energy transmission process remains far from clear. In this issue of Biophysical Journal, a theoretical model is developed to understand the energy transmission process at the axon level near the gray-white matter interface (2).

To investigate the multiscale energy transmission at thresholds that can induce DAI, Alisafaei et al. (2) elegantly combined computational, analytical, and experimental approaches and studied the potential mechanical vulnerability of axons at the gray-white matter junction. Their computational study reveals that material stiffness heterogeneity at the gray-white interface could lead to a tug-of-war mechanism to redistribute axonal strain, which could amplify strain locally. The theoretical prediction was supported by their experimental results from a clinically relevant swine model of concussion that showed the presence, and a notably similar pattern, of axonal damage near the gray-white junction. The gold standard in detecting and evaluating the extent of axonal damage is to observe and quantify the accumulation of amyloid precursor protein that occurs in areas of axonal damage (7). Using amyloid precursor protein immunohistochemistry following a rotational acceleration TBI model at 72 h postinjury, multiple foci of injured axons were observed in their experiments close to the gray-white interface in all injured animals but not in shams. Finally, a simple analytical model was developed to corroborate the suggested mechanism with closely reproduced simulation and experimental results.

These findings provide important insights into where targeted multiscale modeling might be needed. Using a validated head injury model, detailed brain strain distribution can now be efficiently generated along dense white matter fiber tracts (Fig. 1). To understand how impact energy is further transmitted through the meningeal tissue that presents complex biomechanical nature (4) across length scales down to individual axons, a targeted multiscale modeling in specific regions of the brain can be launched, such as the gray-white matter interface. By using macroscale dynamic fiber strain as input, a detailed axonal injury model, such as that in Alisafaei et al. (2), can further reveal the extent of axonal microtubule breakage. Understanding this hierarchical process of energy transmission, brain strain distribution, and redistribution along white matter fibers and axons may hold the key to elucidating the biomechanisms of DAI.

Figure 1.

Detailed fiber strain distribution along dense white matter fiber tracts showing the magnitude of fiber stretch for a concussive head impact at one time point using the anisotropic Worcester Head Injury Model V1.0 (10). To improve visualization, only 10% of a total of ∼35.2 k fiber tracts constructed from ∼3.3 million fiber sampling points are shown. The inset illustrates fiber strain distribution near the gray-white matter junction (shown as a semitransparent surface), which does not yet account for strain redistribution at the axon level. To see this figure in color, go online.

Can any of this be validated experimentally? The pathway forward is likely to be indirect at present. The experimental data that we have for tracking strains in the living human brain arise from magnetic resonance imaging data. However, they are limited to noninjurious, voluntary head motion (8). In addition, they do not yet have the resolution to image at an individual axon level. Nevertheless, using ex vivo histology to validate computational predictions such as in Alisafaei et al. (2) would provide the necessary confidence in model biofidelity. Taken together, Alisafaei et al. (2) show that because of strain redistribution and local amplification at the gray-white matter junction, axonal injury can occur even when the mean tissue strain is below the threshold. This enhances the understanding of the importance of tissue property heterogeneity on the onset of DAI at the axon level (9).

Editor: Guy Genin.