The highways and byways of the brain

Abstract

Brain functions rely on the communication network formed by axonal fibers. However, the number of axons connecting different brain regions is unknown. A study in PLoS Biology addresses this question and finds that most areas of the human cerebral cortex are linked by an astoundingly small number of fibers.

Fig 1. Integrating human brain connectivity at the macro, meso and micro scale.

MR diffusion tractography is a widely used non-invasive method for the macroscopic estimation of anatomical connections in the living human brain (macroscale, top image). Higher resolution, albeit invasive, tractography approaches can only be used in animal model systems or with post-mortem human brain tissue to study projection neurons and axon bundles at the cellular level (mesoscale, bottom left panels of cortical columns in visual cortex labelled with non-phosphorylated intermediate neurofilament protein SMI-32 or with the Gallyas myelin stain). To study individual axons and synaptic interactions between connected neurons and areas one needs to use ultra-high resolution approaches and visualize brain tissue under the electron microscope (microscale, two bottom right panels of myelinated axons in the corpus callosum and deep white matter below the cingulate cortex). By comparing macroscopic counts of MR diffusion streamlines with microscopic counts of axons passing through the corpus callosum, Rosen and Halgren [4] derived a conversion factor that can be applied to tractography data in order to estimate the absolute number of axons in a given tract. However, caution needs to be applied when translating the factor from the well-ordered callosal connections to more heterogeneous non-callosal fiber tracts, such as shown at the bottom right. The calculation suggests that, despite the massive communication system formed by the white matter of the human brain, the absolute number of projections linking any two cortical areas on average is quite small. The finding prompts questions of the modes and neural mechanisms of communication between human cortical areas. Main image: courtesy of the USC Mark and Mary Stevens Neuroimaging and Informatics Institute (www.ini.usc.edu); insets: unpublished data by B. Zikopoulos.