Evidence for corticospinal excitation of presumed propriospinal neurones in man.

1. The possibility that stimulation of the motor cortex facilitates transmission in the pathway mediating non-monosynaptic ('propriospinal') excitation from low-threshold afferents to upper limb motoneurones was investigated. 2. Convergence between peripheral afferent volleys (from the ulnar or musculo-cutaneous nerve) and corticospinal volleys (evoked by magnetic stimulation of the motor cortex) was investigated using the spatial facilitation technique. Thus the effects of these volleys on the flexor carpi radialis H reflex were compared when applied separately and together. When cortical stimulation was optimal for the muscle from which the conditioning volley originated the facilitation of the reflex on combined stimulation was significantly larger than the algebraic sum of the effects of separate stimuli. 3. The extra facilitation on combined stimulation had all the characteristics of 'propriospinal' excitation (low threshold, long central delay, brief duration and depression when the afferent input was increased), and it is suggested that this reflects corticospinal excitation of 'propriospinal' neurones. 4. When varying the time interval between cortical and test stimulations, it was shown that extra facilitation on combined stimulation began 1 ms later than the onset of the control reflex facilitation. Assuming that the latter onset reflects the arrival of the monosynaptic corticospinal volley at the motoneurone pool, this 1 ms delay suggests a disynaptic pathway for the cortical excitation of motoneurones through 'propriospinal' neurones. 5. As at the onset of voluntary movement, the pattern of the cortical excitation of 'propriospinal' neurones was quite specific: extra facilitation of the reflex on combined stimulation only occurred when the cortical volley was preferentially directed to motoneurones supplying the muscle from which the afferents used for the peripheral volley originated. 6. It is concluded that corticospinal axons activate human 'propriospinal' neurones and thereby produce disynaptic excitation of the motoneurone pool. Given temporal summation with the monosynaptic excitation, this 'propriospinally mediated' disynaptic excitation might make a significant contribution to the evoked EMG potential.