Correlations Between Corticomotoneuronal (CM) Cell Postspike Effects and Cell-Target Muscle Covariation

Abstract

The presence of postspike facilitation (PSpF) in spike-triggered averages of electromyographic (EMG) activity provides a useful means of identifying cortical neurons with excitatory synaptic linkages to motoneurons. Similarly the presence of postspike suppression (PSpS) suggests the presence of underlying inhibitory synaptic linkages. The question we have addressed in this study concerns the extent to which the presence and strength of PSpF and PSpS from corticomotoneuronal (CM) cells correlates with the magnitude of covariation in activity of the CM cell and its target muscles. For this purpose, we have isolated cells during a reach and prehension task during which the activity of 24 individual proximal and distal forelimb muscles was recorded. These muscles show broad coactivation but with a highly fractionated and muscle specific fine structure of peaks and valleys. Covariation was assessed by computing long-term (2 s) cross-correlations between CM cells and forelimb muscles. The magnitude of cross-correlations was greater for muscles with facilitation effects than muscles lacking effects in spike-triggered averages. The results also demonstrate a significant relationship between the sign of the postspike effect (facilitation or suppression) and the presence of a peak or trough in the cross-correlation. Of all the target muscles with facilitation effects in spike-triggered averages (PSpF, PSpF with synchrony, or synchrony facilitation alone), 89.5% were associated with significant cross-correlation peaks, indicating positively covarying muscle and CM cell activity. Seven percent of facilitation effects were not associated with a significant effect in the cross-correlation, whereas only 3.4% of effects were associated with correlation troughs. In contrast, of all the muscles with suppression effects in spike-triggered averages, 38.9% were associated with significant troughs in the cross-correlation, indicating an inverse relation between CM cell and muscle activity consistent with the presence of suppression. Fifty-five percent of suppression effects was associated with correlation peaks, whereas 5.6% was not associated with a significant effect in the cross-correlation. Limiting the analysis to moderate and strong facilitation effects, the magnitude of PSpF was correlated weakly with the magnitude of the cell-muscle cross-correlation peak. Nevertheless, the results show that although many CM cell-target muscle pairs covary during the reach and prehension task in a way consistent with the sign and strength of the CM cell's synaptic effects on target motoneurons, many exceptions exist. The results are compatible with a model in which control of particular motoneuron pools reflects not only the summation of signals from many CM cells but also signals from additional descending, sensory afferent, and intrinsic spinal cord neurons. Any one neuron will make only a small contribution to the overall activity of the motoneuron pool. In view of this, it is not surprising that relationships between postspike effects and ...

Prediction of Muscle Activity by Populations of Sequentially Recorded Primary Motor Cortex Neurons

“…Others have examined the correlation or coherence between EMG signals and either the discharge of single neurons (Holdefer and Miller 2002;McKiernan et al 2000;Schwartz and Adams 1995), local field potentials (Baker et al 1999), or electroencephalographic signals (Halliday et al 1998;Mima et al 2000). However, to the best of our knowledge, there has been only one other study that attempted to account for the time course of EMG modulation on the basis of the discharge of multiple neurons (Fetz et al 1989).…”

Joint cross-correlation analysis reveals complex, time-dependent functional relationship between cortical neurons and arm electromyograms

“…Additionally, lag-dependent correlations may occur through indirect effects. For example, this effect could emerge if a given cortical motoneuron does not innervate the muscle being recorded, but instead innervates a different muscle that acts in synchrony with the former muscle (McKiernan et al 2000). It is also possible that some of the lag-dependent correlations observed in our study reflected the cortical unit's responses to proprioceptive feedback generated by muscle contractions.…”

“…Several previous studies came to similar conclusions. McKiernan et al (2000) reported only a weak relationship between short-latency postspike effects (i.e., a relatively stable relationship based on direct connectivity) and more broad neuron-muscle covariations. Another study (Schwartz and Adams 1995) reported time-dependent changes in cortico-EMG cross-correlation during a sinusoidal tracking task.…”

“…The now classic STA approach, which effectively represents cross-correlation between single neurons and EMGs, was introduced to assess functional connectivity of spinal motoneurons Jackson et al 2006;Mantel and Lemon 1987;McKiernan et al 2000;Miller et al 1993;Morrow and Miller 2003;Schieber and Rivlis 2007). This and similar methods have been used to analyze the inputs to the spinal cord from the cortex (Carmena et al 2003;Davidson et al 2007;Fuglevand 2011;McKiernan et al 2000;Morrow et al 2007;Schaefer et al 2009;Schieber 2011), red nucleus (Miller et al 1993) and reticular formation (Stuphorn et al 1999). The analytic approach employed in the present study, JCC, is very similar to the JPSTH method introduced in the late 1960s for the analysis of time-dependent correlations between neurons (Aertsen et al 1989;Cardoso de Oliveira et al 2001;Gerstein and Perkel 1969;Vaadia et al 1995).…”

“…This approach has proven to be useful to detect short-latency mono-and polysynaptic connections between cortical neurons and spinal motoneurons during different motor behaviors (Baker et al 1997;Buys et al 1986;Davidson et al 2007;Jackson et al 2003;McKiernan et al 2000;Muir and Lemon 1983;Smith and Fetz 1989). STA usually represents average comodulations taken from many different behavioral epochs (McKiernan et al 1998;Semmler et al 2000), even though STA variations between different epochs of behavioral tasks (Baker et al 1997;Morrow and Miller 2003) and different tasks (Zhou and Fuster 2000) have also been reported.…”