Stereotypical Bouton Clustering of Individual Neurons in Cat Primary Visual Cortex

Abstract

In all species examined, with the exception of rodents, the axons of neocortical neurons form boutons in multiple separate clusters. Most descriptions of clusters are anecdotal, so here we developed an objective method for identifying clusters. We applied a mean-shift cluster-algorithm to three-dimensional reconstructions of 39 individual neurons and three thalamic afferents from the cat primary visual cortex. Both spiny (20 of 26) and smooth (7 of 13) neurons formed at least two distinct ellipsoidal clusters (range, 2-7). For all cell types, cluster formation is heterogenous, but is regulated so that cluster size and the number of boutons allocated to a cluster equalize with increasing number of clusters formed by a neuron. The bouton density within a cluster is inversely related to the spatial scale of the axon, resulting in a four times greater density for smooth neurons than for spiny neurons. Thus, the inhibitory action of the smooth neurons is much more concentrated and focal than the excitatory action of spiny neurons. The cluster with the highest number of boutons (primary cluster) was typically located around or above the soma of the parent neuron. The distance to the next cluster was proportional to the diameter of the primary cluster, suggesting that there is an optimal distance and spatial focus of the lateral influence of a neuron. The lateral spread of clustered axons may thus support a spoke-like network architecture that routes signals to localized sites, thereby reducing signal correlation and redundancy.

Anatomical Constraints on Lateral Competition in Columnar Cortical Architectures

“…We gratefully acknowledge Tom Binzegger, Kevan Martin and Rodney Douglas for providing the data in Figure 1 (Binzegger et al, 2007). We thank Rodney Douglas for spurring this work on in its early stages.…”

“…(B) The effective lateral coupling for a slice through layer 2/3 of cat visual cortex (Binzegger, Douglas, & Martin, 2007). Strong local inhibition coupled with long-range excitation results in a net profile with local inhibitory dominance and much weaker lateral excitatory dominance.…”

“…Inhibitory synapses were estimated to be 10 times stronger than excitatory synapses (Binzegger, Douglas, & Martin, 2009). (C) Raw synaptic profiles of excitation (positive curves) and inhibition (negative curves) for layer 2/3 of cat visual cortex (Binzegger et al, 2007). Faint curves show data from individual reconstructed neurons; darker curves indicate the average over the set of reconstructed neurons.…”

Eigenspectrum bounds for semirandom matrices with modular and spatial structure for neural networks

“…Random matrices that incorporate this constraint are known as Euclidean random matrices [28], and have been used to model physical phenomena such as diffusion [29] and wave propagation [30]. Neurons in the neocortex reside in physical space, with connection probabilities modulated smoothly across the cortical surface [31][32][33]. In addition, connections within the cortex are modulated by functional similarity over one or more physiological metrics.…”

Fundamentals of Natural Representation

“…Considering how the conjunction of correlation profiles limits the range of positive correlation sharply around the specifics of reality present in the observed system, it is an advantage to a processing system to capture or sample as much of a diversity in the correlation profiles as possible with the agents [24,48,49] within the sampling domains. Even the feedback profiles could initially bear greater diversity.…”