A numerical analysis of the geniculocortical input to striate cortex in the monkey

Using data that are available in various publications, a quantitative analysis has been made of the geniculocortical input to layer IVC of the macaque striate cortex. The data suggest that only 1.3-1.9% of the excitatory, or asymmetric synapses in layer IVC alpha of striate cortex are provided by the neurons of the magnocellular layers of the LGN. This amounts to only 18-40 of the 1000-2100 asymmetric synapses that the average layer IVC alpha neuron receives. The parvicellular afferents to layer IVC beta, on the other hand, provide 3.7-8.7% of the asymmetric synapses formed by the average layer IVC beta neuron, or 37-191 synapses to each neuron. If it is assumed that the boutons in the geniculocortical axonal plexuses are evenly spread, it can be calculated that the input to an individual layer IVC neuron is provided by some 24 axonal plexuses. This is regardless of whether the neuron lies in layer IVC alpha or in IVC beta. This calculation suggests that a single axonal plexus provides not more than one or two of the excitatory synapses received by an individual layer IVC alpha neuron, and between one and eight excitatory synapses for a layer IVC beta neuron. Consequently, it is unlikely that the response properties of a particular cortical neuron are dominated by its input from a single geniculate neuron. Since the geniculocortical input essentially determines the response properties of neurons in layer IV of macaque striate cortex, it is surprising that this input amounts to such a small number of synapses to an individual neuron, although we obtained a somewhat similar result in our earlier quantitative analysis of the geniculate input to the striate cortex of the cat (Peters and Payne, 1993). But it has to be questioned whether the low values obtained are correct. Interestingly, the geniculocortical input to cortex has been largely neglected in favor of analyses of intracortical circuitry, but in view of the basic importance of this afferent input, it is suggested that more quantitative data about it should be generated, so that a better assessment can be made of its extent.