Representation of facial muscles in human motor cortex

Abstract

Whether there is a projection from the primary motor cortex (M1) to upper facial muscles and how the facial M1 area is modulated by intracortical inhibitory and facilitatory circuits remains controversial. To assess these issues, we applied transcranial magnetic stimulation (TMS) to the M1 and recorded from resting and active contralateral (C-OOc) and ipsilateral orbicularis oculi (I-OOc), and contralateral (C-Tr) and ipsilateral triangularis (I-Tr) muscles in 12 volunteers. In five subjects, the effects of stimulating at different scalp positions were assessed. Paired TMS at interstimulus intervals (ISIs) of 2 ms were used to elicit short interval intracortical inhibition (SICI) and ISI of 10 ms for intracortical facilitation (ICF). Long interval intracortical inhibition (LICI) was evaluated at ISIs between 50 and 200 ms, both at rest and during muscle activation. The silent period (SP) was also determined. C-OOc and I-OOc responses were recorded in all subjects. The optimal position for eliciting C-OOc responses was lateral to the hand representation in all subjects and MEP amplitude markedly diminished when the coil was placed 2 cm away from the optimal position. For the I-OOc, responses were present in more scalp sites and the latency decreased with more anterior placement of the coil. C-Tr response was recorded in 10 out of 12 subjects and the I-Tr muscle showed either no response or low amplitude response, probably due to volume conduction. SICI and ICF were present in the C-OOc and C-Tr, but not in the I-OOc muscle. Muscle activation attenuated SICI and ICF. LICI at rest showed facilitation at 50 ms ISI in all muscles, but there was no significant inhibition at other ISIs. There was no significant inhibition or facilitation with the LICI protocol during muscle contraction. The SP was present in the C-OOc, C-Tr and I-OOc muscles and the mean durations ranged from 92 to 104 ms. These findings suggest that the I-OOc muscle response is probably related to the first component (R1) of the blink reflex. There is M1 projection to the contralateral upper and lower facial muscles in humans and the facial M1 area is susceptible to cortical inhibition and facilitation, similar to limb muscles.

Figure 1. Positions of the coil with respect to the optimal position for eliciting MEPs from the contralateral orbicularis oculi muscle (C-OOc, position 0) in the anterior, posterior, lateral and medial directions

The numbers in each position denote the distance (in cm) from the optimal position. The hand represents the optimal position for eliciting MEPs from the contralateral first dorsal interosseus muscle. In most subjects, it is about 2 cm medial and 1 cm posterior to the optimal position for the C-OOc muscle.

Figure 2. Responses from the right (R) and left (L) orbicularis oculi (OOc), frontalis (Fr), triangularis (Tr) and masseter (Ma) muscles with the coil placed over the left M1 at the optimal position for eliciting a response from the R OOc muscle

Stimulus intensity was 90% of the maximal stimulator output. Four trials from one subject were superimposed. All muscles were at rest in the left column, whereas the responses of the right column were recorded during bilateral OOc muscle contraction. The arrowheads and the numbers below the R and L OOc and the L Ma responses denote the onset latencies. OOc activation results in increased amplitude and decreased latency of the R OOc MEPs. There were no changes in the amplitude, shape or latency in the L OOc and L Ma responses, suggesting R1 of the blink reflex followed by R2 for L OOc and compound muscle action potential due to stimulation of the motor axons of the trigeminal nerve for L Ma.

Figure 3. Responses from the right (R) and left (L) orbicularis oculi (OOc), frontalis (Fr), triangularis (Tr) and masseter (Ma) muscles with the coil placed in the optimal position for eliciting a response from the R OOc muscle

Four trials from one subject were superimposed. Stimulus intensity was 67% of the maximal stimulator output. All muscles were at rest in the left column, whereas the responses of the middle column were recorded during bilateral OOc muscle contraction and those of the right column during bilateral Tr muscle contraction. The arrowheads and the numbers below denote the onset latencies. OOc activation results in increased R OOc MEP amplitude. There is also a moderate amplitude increase in the R Fr and R Tr muscles. Tr muscle activation results in marked facilitation of the R Tr response, which may be transmitted via volume conduction to the other muscles recorded.

Figure 4. Responses of the contralateral and ipsilateral orbicularis oculi (OOc) muscles at rest, with TMS at different scalp sites

For comparison, A shows the blink reflex elicited with the centre of the coil applied over the supraorbital notch (3 superimposed trials). The number below the trace of the I-OOc denotes the onset of the early response of the blink reflex (R1). There is no R1 in the contralateral OOc. The arrows indicate the onset of the late response of the blink reflex (R2) at 35.6 ms in the ipsilateral and 35.9 ms in the contralateral OOc. Responses in the anterior–posterior axis are shown in B, and with TMS at different scalp sites in the coronal plane in C. Each trace represents 5 superimposed trials in one subject. Stimulus intensity was 70% of the maximal stimulator output in all sites. The numbers preceding the traces denote the distance (in cm) with respect to the optimal stimulation site, with anterior being positive and posterior negative in B, and with medial being positive and lateral negative in C. Zero represents the optimal position for the contralateral OOc response. The numbers below the responses denote the onset latencies (in ms). TMS elicited a maximal response in the contralateral OOc muscle when the coil was placed in the optimal site. When the coil was displaced 2 cm in any direction, there was a pronounced reduction in amplitude and further displacement resulted in no response. In B, the ipsilateral OOc muscle showed responses of similar shape and amplitude between 2 cm posterior to the optimal site and 6 cm anterior to the optimal site. The onset latency decreased with progressively more anterior placement of the coil, eventually approaching the latency of the early response of the BR. In C, the ipsilateral OOc muscle showed a response with similar shape and amplitude with the coil placed 2 cm medial or lateral to optimal site. Stimulation at 4 and 6 cm lateral to the optimal site produced a large, short latency response due to direct facial nerve stimulation.

Figure 5. Direct stimulation of the right and left facial nerves and responses with the coil placed 3 cm above the scalp (acoustic effect)

Eyes were open and facial muscles were at rest. Each trace shows three superimposed trials. The vertical line shows the onset of late responses (LR), which occurred at latencies of about 35 ms in all traces. Both types of stimulation resulted in bilateral LR. The latency of the OOc M response was 3.0 ms on both sides and the amplitude was 1.2 mV on the right side and 1.5 mV on the left side. R, right; L, left; Ooc, orbicularis oculi; M, M wave; LR, late response.

Figure 6. Silent period (SP) of contralateral (C) and ipsilateral (I) orbicularis oculi (OOc) muscles in one subject

Twenty sweeps were superimposed. The horizontal bars indicate the duration of the silent period measured from the stimulus artifact to the first return of EMG activity. Both C- and I-OOC show a consistent SP following the motor-evoked potential (MEP) which is interrupted by a long latency reflex in the latency range of R2 of the blink reflex.

Figure 7. SICI and ICF at rest

The figure shows recordings from the contralateral orbicularis oculi (C-OOc), contralateral triangularis (C-Tr) and ipsilateral orbicularis oculi (I-OOc) in one subject, with the conditioning stimulation intensity set at 80% of the resting motor threshold. Ten sweeps have been superimposed in each trace. The left column shows responses with test stimulation alone (Test), the middle column with 2 ms and right column with 10 ms interstimulus interval (ISI). C-OOc and C-Tr show decreased amplitude at 2 ms ISI (SICI), and increased amplitude at 10 ms ISI (ICF) compared to test stimulation alone. There was no significant change in MEP amplitude in the I-OOc muscle.

Figure 8. Grouped data for SICI and ICF values for rest and during muscle activation

SICI and ICF for contralateral orbicularis oculi (C-OO), contralateral triangularis (C-Tr) and ipsilateral orbicularis oculi (I-OO) with conditioning stimulation intensities set at 80% (Rest 80, open columns) and 95% (Rest 95, filled columns) of the motor threshold, and during muscle activation (Active, hatched columns). The amplitude of the motor-evoked potential (MEP) is expressed as a ratio to the MEP of test stimulus alone (vertical axis). Values above one indicate or facilitation and values below one indicate inhibition. Error bars represent s.e.m.* Significant inhibition facilitation at P < 0.017. The C-OOc muscle showed SICI₈₀, SICI₉₅ and active SICI. The C-Tr muscle had SICI₈₀ and SICI₉₅, but muscle activation resulted in no SICI. There was no significant inhibition in the I-OOc muscle. At rest, there was significant ICF in C-OO and C-Tr.

Figure 9. LICI at rest (left) and during activation (right)

The amplitude of the motor-evoked potential (MEP) was expressed as a ratio to that of the test stimulus alone (vertical axis) in the contralateral orbicularis oculi (C-OOc, open columns), in the contralateral triangularis (C-Tr, filled columns) and in the ipsilateral orbicularis oculi (I-OOc, hatched columns) muscles at ISIs of 50, 100, 150 and 200 ms. Values above one indicate facilitation and values below one indicate inhibition. Error bars represent s.e.m.* Significant inhibition or facilitation at P < 0.01. At rest there was significant facilitation at 50 ms ISI in all three muscles. There was no significant inhibition.