BIO254:Columns

OVERVIEW
Neurons in many cortical areas are arranged into functional columnar structures spanning from the pial surface to the white matter tracts. A cortical column is defined by a group of neurons arranged vertically that share a similar receptive field. For example, as an electrode oriented perpendicular to the surface of the primary visual cortex is penetrated deeper into the cortex, all neurons encountered will respond to a bar of light angled at 45 degrees from the horizon (Figure 1, left)1. However, neurons recorded from an electrode inserted parallel to the cortical surface will show gradually changing orientation selectivity (Figure 1, right). Columnar organization was first observed in the somatosensory cortex by V.B. Mountcastle in 19572. Since then columnar organization has been described in the visual cortex3, auditory cortex4,5, and motor cortex6.



Initial Work
Vernon Mountcastle was the first to rigoursly describe functional organization of neurons into cortical columns2. Using a microelectrode, Montcastle recorded from cat primary somatosensory cortex and characterized neurons with receptive fields to one of three stimuli: movement of hair, cutaneous pressure, or deep pressure. Further, he found that neurons responding to the same mechanostimulation to the same area of the body were arranged vertically into columns. Following Mountcastle’s discoveries, David Hubel and Torsten Wiesel found orientation columns in cat and monkey visual cortex by using electrophysiology and 14C 2-deoxyglucose, a metabolic marker of active neurons (Figure 2). From these experiments, Hubel and Wiesel suggested that the orientation columns formed maps consisting of "swirling stripes with many bifurcations and endings7." The data from these experiments, along with many subsequent studies, was analyzed by V. Braitenberg and C. Braitenberg8. From their analysis, the notion of circularly symmetric orientations around central points was suggested, which later developed into the concept of pinwheels.

Non-Invasive Optical Imaging
Limitations of electrophysiology and 2-deoxyglucose staining prevented precise mapping of orientation columns over large areas of cortex. It was not until the mid-1980's that Gary Blasdel and Guy Salama provided a new method to study cortical columns in vivo9. By utilizing voltage-sensitive dyes, Blasdel and Salama concluded that orientation preference changes gradually within domains and that these domains overlie ocular dominance columns. Using the voltage sensitive dye technique developed by Blasdel and Salama, Tobias Bonhoeffer and Amiram Grinvald were able to show visually the existence of pinwheel structures10, supporting the hypothesis put forth by Braitenberg. In addition, they were able to correlate sudden jumps in orientation preference seen in electrophysiological recordings to the centers of the pinwheels.



Single Cell Imaging
Recently, Ohki et al. used two-photon microscopy with calcium sensitive dyes to map for the first time complete functional areas at single neuron resolution6. Supporting previous studies, Ohki et al. first showed that there is no functional organization to rat visual cortex. They next moved to the highly organized cat visual cortex and demonstrated that in a given domain, orientation preference changes smoothly in space. However, in some areas, orientation preference can abruptly switch with boundaries between columns being only a couple of cells wide (Figure 4). They subsequently demonstrated that even neurons in the pinwheel centers are highly organized and orientation selective (Figure 5)12. A major question arises from the discrepancy between functional organization in the cat and lack of organization in the rat: What is the significance of cortical columns if they are not necessary for visual processing?