The role of the γ-system for improving information transmission in populations of Ia afferents

Ensemble coding of simple mechanical stimuli (small sinusoidal stretches) was studied in populations of simultaneously recorded primary muscle spindle afferents (MSAs). The experiments were made on 39 primary MSAs in choralose anaesthetised cats. For the analyses we used a combination of principal component analysis and algorithms for quantification of stimulus discrimination. Ensembles of primary MSAs discriminated better between different muscle stretches than individuals, and showed a successive increase in discriminative ability with increasing ensemble size. The ensemble effect disappeared after cutting the ventral roots, indicating an important role for the fusimotor system in ensemble coding. Simultaneously recorded ensembles of MSAs showed significantly better discriminative ability than pooled sequentially recorded MSAs. This difference was abolished by the cutting of the ventral roots. It is hypothesised that, since the muscle spindles are connected to each other via secondary MSAs (projecting to γ-motoneurones to other spindles), the fusimotor-muscle spindle system may constitute a neural network, in which each ‘neuron’ (i.e., each individual muscle spindle) is influenced by the activity in the whole network. In populations of pooled sequentially recorded MSAs, the connections would not exist. Thus, the population would lose its neural network quality, and the encoding ability of the population would decrease.