Brain, Vol. 126, No. 1, 115-133,
January 2003
© 2003 Guarantors of Brain
doi: 10.1093/brain/awg008
Motoneuron adaptability to new motor tasks following two types of facial–facial anastomosis in cats
1 Laboratorio Andaluz de Biología, Universidad Pablo de Olavide, Sevilla, Spain, 2 Klinik für Hals-, Nasen- und Ohrenheilkunde and 3 Institut I für Anatomie der Universität zu Köln, Köln, Germany
Correspondence to: Professor José M. Delgado-García, MD, PhD, Laboratorio Andaluz de Biología, Universidad Pablo de Olavide, Ctra de Utrera, Km. 1, 41013-Sevilla, Spain E-mail: jmdelgar{at}dex.upo.es
Received April 22, 2002. Revised July 21, 2002. Accepted August 9, 2002.
 |
Summary |
|---|
 Top Summary IntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The ability of the facial motor system to adapt to a new motor function was studied in alert cats after unilateral transection, 180° rotation and suture of the zygomatic nerve, or transection and cross-anastomosis of the proximal stump of the buccal nerve to the distal stump of the zygomatic nerve. These procedures induced reinnervation of the orbicularis oculi (OO) muscle by different OO- or mouth-related facial motoneurons. Eyelid movements and the electromyographic activity of the OO muscle were recorded up to 1 year following the two types of anastomosis. Animals with a zygomatic nerve rotation recovered spontaneous and reflex responses, but with evident deficits in eyelid kinematics, i.e. the proper regional distribution of OO motor units was disorganized by zygomatic nerve rotation and resuture, producing a permanent defect in eyelid motor performance. Following buccal–zygomatic anastomosis, the electrical activity of the OO muscle was recovered after 6–7 weeks, but air puff-, flash- and tone-evoked reflex blinks never reached the control values on the operated side. Electromyographic OO activities and lid movements corresponding to licking and deglutition activities were observed on the operated side in buccal–zygomatic anastomosed animals up to 1 year following surgery. Mouth-related facial motoneurons did not readapt their discharges to the kinetic, timing and oscillatory properties of OO muscle fibres. A significant hyper-reflexia was observed following both types of nerve repair in response to air puffs, but not to light flashes or tones. In conclusion, adult mammal facial premotor circuits maintain their motor programmes when motoneurons are induced to reinnervate a foreign muscle, or even a new set of muscle fibres.
Keywords: blinks; corneal reflex; facial motor system; neural plasticity; reinnervation
Abbreviations: EMG= electromyography; HRP = horseradish peroxidase; OO = orbicularis oculi
|
Introduction |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Although it is a general assumption that neural centres and circuits are intrinsically able to modify their firing activity in order to re-specify their commands to different environmental constraints, the place where these changes take place in the nervous system and how these modifications are achieved are matters of continuing debate (Woody, 1986
; Dobkin, 1997; Kim and Thompson, 1997). Since motoneurons represent the final common pathway where motor commands are executed, it would be interesting to clarify to what degree such adaptive behaviour can occur in motor nuclei and/or in their premotor circuits in adult mammals. Sperry (1945) stated clearly that adult mammal neural motor centres lack the capacity to reorganize their discharge properties to fulfil the functional requirements of new motor targets when involved motoneurons de novo innervate a different muscle. Nevertheless, the issue has been reopened from time to time as a result of both clinical reports and animal experiments (Stennert, 1979; Gruart et al., 1995b; Willer et al., 2002).
Indeed, the available data on the facial motor system are somewhat controversial. Hypoglossal–facial anastomosis today is a standard method for the treatment of some types of facial paralysis in humans (Körte, 1903; Stennert, 1979; May et al,., 1991). Besides the prevention of facial muscle atrophy and the subsequent asymmetry in facial expression, some changes in motor function following reinnervation of the facial musculature by hypoglossal motoneurons have been described, including the restored ability for some spontaneous smiling (Stennert, 1979; Miehlke et al., 1981; Hammerschlag, 1987; May et al., 1991; Tankéré et al., 2000). Even a possible rerouting of central sensorimotor pathways has been proposed to support these functional changes (Willer et al., 2002). In contrast, a report regarding functional changes in eyelid motor responses following hypoglossal–facial anastomosis in the cat indicates that hypoglossal motoneurons are able to survive while innervating the orbicularis oculi (OO) muscle, but their firing codes are not re-specified or adapted to the kinematics and frequency-domain properties of the eyelid motor system (Gruart et al., 1996; Domingo et al., 1997). Moreover, available data from the extraocular motor system of adult cats indicate that most ocular motoneurons are able to survive the transection of the corresponding peripheral oculomotor cranial nerve, but lack the capacity for finding their parent muscle and, more importantly, their firing rates are unrelated to the functional needs of the reinnervated muscle (Baker et al., 1985; Baker, 1986).
The present experiments were designed to study both spontaneous and reflex motor eyelid responses in adult cats following (i) ipsilateral zygomatic nerve section, 180° rotation and resuture; or (ii) buccal–zygomatic anastomosis, i.e. transection and crossed suture of the proximal buccal stump to the distal zygomatic stump of these two branches of the facial nerve. With these two experimental procedures, we attempted to enforce the possible central reorganization of sensorimotor pathways and/or specific commands to motoneurons of different function, but located within the same nucleus. Thus, either OO- or mouth-related, motoneuron axons were redirected to motor units originally innervated by different OO motoneurons. We decided to study the eyelid motor system because of the technical facilities available for its observation, recording and analysis both in humans (Kilimov and Linke, 1978; Stennert and Limberg, 1982) and in cats (Gruart et al., 1995a, 1996; Domingo et al., 1997). The kinematics and frequency-domain properties of air puff-, flash- and tone-evoked reflex blinks were studied in alert cats up to 1 year following these two types of facial–facial anastomosis. Eyelid responses characterizing cat’s friendly displays and winking (Gruart et al., 1995a), or those accompanying licking movements, were also recorded and analysed. Eyelid movements and the electromyographic (EMG) activity of the OO muscle were recorded bilaterally and analysed. The location of motoneurons reinnervating the OO muscle was checked by histological procedures.
|
Material and methods |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Subjects
Nine adult female cats, weighing from 2.1 to 2.6 kg, were used in the present experiments. Animals were obtained from an authorized supplier (Iffa-Credo, France). All experimental procedures were carried out in accordance with the guidelines of the European Union Council (86/609/EU) and following the Spanish regulations (BOE 67/8509-12, 1988) for the use of laboratory animals in chronic experiments. The nine animals were implanted bilaterally with search coils and muscle electrodes for the chronic recording of upper eyelid movements and of the EMG activity of the OO muscle. A transection, 180° rotation of the proximal stump and resuture of the left zygomatic nerve was performed in three of the animals. In addition, a buccal–zygomatic cross-anastomosis, also on the left side of the face, was carried out in another three animals. The remaining three animals served as controls.
Surgical procedures
Animals were anaesthetized with sodium pentobarbital (Sigma, MO, USA, 35 mg/kg i.p.) after a protective injection of atropine sulfate (0.5 mg/kg i.m.) to prevent unwanted vagal reflexes. A five-turn coil (3 mm in diameter) was implanted bilaterally into the centre of the upper eyelid at 
2 mm from the lid margin. Coils were made of seven-strand, Teflon-coated stainless steel wire, with an external diameter of 50 µm (A-M Systems, WA, USA). Because of their low weight (10–15 mg), coils did not impair movement or cause any drooping of the lids. Animals were also implanted with bipolar hook electrodes in both OO muscles. These EMG-recoding electrodes were made of the same wire as the lid coils, and 1 mm of their tip was bared of Teflon. A 1 mm diameter, non-isolated silver electrode was attached to the skull as ground. Finally, the animals were implanted with a head-holding system for stabilization during recording sessions. Three bolts were fixed to the skull with dental cement, perpendicularly to the horizontal stereotaxic plane. Eyelid coils, and EMG and ground electrodes were soldered to a nine-pin socket attached to the holding system.
One month later, six of the animals were re-anaesthetized with the above-described procedure. In three of the animals, the left zygomatic nerve was exposed and transected. The central stump of the nerve was rotated 180° and sutured end-to-end to the distal stump with three 11-0 atraumatic sutures (Ethilon EH 7438G, Ethicon, Germany). In another three animals, the left facial nerve was exposed, and both the buccal and zygomatic branches were transected. The central stump of the buccal branch was then sutured end-to-end to the distal stump of the zygomatic nerve with the same atraumatic sutures. The proximal stump of the zygomatic branch of the facial nerve was displaced caudally and sutured to the nearby ear muscles (see Fig. 1A).
|
General conditions of recording sessions
Control recording sessions started 5 days following implantation surgery and, unless otherwise stated, lasted a maximum of 2 h/day. Control values (in Figs 3, 6, 8 and 9) were recorded from the nine animals 20 days following implantation surgery, i.e. 10 days prior to crossed nerve suture. Recording sessions were restarted 5 days following the two types of facial–facial anastomosis (month 0, in Figs 3, 6, 8 and 9) and were repeated 3, 6, 9 and 12 months later. Brief (
20 min) recording sessions were carried out weekly following the two types of anastomosis to determine the time of reinnervation of the OO muscle. For recordings, the animal was lightly restrained with an elastic bandage, placed on the recording table, and its head immobilized by attaching the head-holding system to a bar fixed to the table. The animals seemed not to be stressed by the experimental procedures, since the monitoring of their cardiac and respiratory rhythms with non-invasive methods throughout the recording sessions yielded values similar to those obtained with the animals resting in the arms of one of the experimenters (Gruart et al., 1995a).
|
|
|
|
Recording and stimulating techniques
Lid movements were recorded with the search coil in a magnetic field technique (Gruart et al., 1995
a). Each coil was calibrated with the cat still under the effects of the anaesthetic, by rotating the magnetic frame ±10° in the vertical plane and measuring the output voltage of the coil while displacing the lid manually at selected angles. The EMG activity of OO muscles was recorded with differential amplifiers between 1 Hz and 10 kHz. The animal was presented with stimuli of different sensory modalities (puffs of air, flashes of light and tones) aimed at evoking blinking responses. These stimuli were presented randomly, at 30 ± 10 s intervals, at least 20 times/session to both control and experimental sides. Air puffs were applied through a pipette located 1 cm away from the cornea, at a pressure of 3 kg/cm2, and lasting 100 ms. Flashes of light were provided by a xenon arc lamp located 1 m in front of the animal’s eyes, and lasted 1 ms. Tones (600 and 6000 Hz) were delivered for 100 ms at 90 dB. The loudspeaker was located 1 m below the animal’s head. Licking was evoked in the animals by the application, with a home-made dispenser, of a few milk drops at the tongue tip. Friendly responses, spontaneous blinks and winking responses were made by the animals randomly across recording sessions (see Fig. 1B).
Histology
At the end of the recording sessions, the six animals with anastomosis and the three controls were anaesthetized as described above and 2 mg of horseradish peroxidase (HRP, grade I; Boehringer Mannheim, Germany), dissolved in 0.2 ml of distilled water containing 2% DMSO (dimethylsulfoxide), was injected under the skin of the external canthus, close to the left zygomatic nerve. Two days later, the animals were deeply re-anaesthetized (sodium pentobarbital, 50 mg/kg i.p.) and perfused with saline followed by a mixture of 1.25% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M phosphate buffer pH 7.4. The brain was removed, the anastomosed side was marked and the brainstem was cut in 50 µm coronal serial sections with a vibratome. The HRP enzymatic activity was revealed with the tetramethylbenzidine procedure (Mesulam, 1978) in all sections through the facial nucleus to identify labelled motoneurons. The number of labelled motoneurons within the different subdivisions of the facial nucleus was determined, and their location charted, for both control and anastomosed animals following procedures described elsewhere (Neiss et al., 1992; Angelov et al., 1993; Guntinas-Lichius et al., 1993, 1996). The proper location of the recording electrodes was checked in the nine animals.
Analysis of the recorded data
Vertical and horizontal position of both eyelids, unrectified EMG activity of both OO muscles, and trigger pulses corresponding to blink-evoking stimuli were stored digitally on an eight-channel video tape system. Data were transferred later to a computer, through a CED 1401-plus converter, at 1–4 kHz sampling frequency, with an amplitude resolution of 12 bits. Computer programs were designed to represent lid position, velocity and acceleration, and the EMG activity of the OO muscle. The programs also allowed the measurement of latency, amplitude, duration, velocity and acceleration of evoked movements, and of latency, and peak and integrated EMG activity of the OO muscle. The integrated EMG amplitude was measured after its rectification. Velocity and acceleration traces were computed digitally as the first and second derivative of lid position traces, following low-pass filtering of the data (–3 dB cutoff at 50 Hz and a zero gain at 100 Hz; see Domingo et al., 1997).
Statistical analyses were carried out with the SPSS/PC+ package (SPSS Inc, IL, USA). Mean values for lid movement latencies, amplitudes, velocities and accelerations, as well as for EMG latency, and peak and integrated amplitudes, were calculated from 
15 measurements collected from the anastomosed or non-anastomosed animals. Mean values are presented, followed by their standard deviation. The quantitative differences between measurements obtained from control (right) and anastomosed (left) sides were compared through the 12 months that the study lasted, using a two-way (side by time) analysis of variance (MANOVA) with repeated measurements in the second factor at a significance level of P = 0.01. A one-way repeated measurement ANOVA was used for data collected from the control side for those parameters that did not present any noticeable value in the anastomosed side. The polynomial contrast test from MANOVA was used to assess any trend in reflexively evoked blinks during the 1 year of this study.
The power of the spectral density function of eye acceleration recordings was calculated using the fast Fourier transform to determine the relative strength of the different frequencies present in lid responses (for details see Bendat and Piersol, 1986; Wessberg and Vallbo, 1995; Domingo et al., 1997). Briefly, acceleration recordings corresponding to lid responses were divided into 1.024 s segments, starting 100 ms before the presentation of the reflex stimulus. Illustrated power spectra (Fig. 2) correspond to the mean value of power spectra computed from 15 different acceleration segments. Peaks of power spectra were tested with the 
2-distributed test for spectral density functions.
|
|
Results |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
General observations
After the two types of nerve suture and before reinnervation, animals showed no blinking of the left eyelid in response to mild blows of air directed to the ipsilateral cornea. Reinnervation of the right OO muscle took place 6–7 weeks following zygomatic nerve rotation and buccal–zygomatic anastomosis, as indicated by the presence of some spontaneous background activity in the EMG recordings. In the case of the buccal–zygomatic anastomosis, the recovered EMG activity was related mainly to buccal movements, such as during licking and mastication. It was quite remarkable that buccal–zygomatic anastomosed animals moved their left eyelids overtly while eating or drinking. Zygomatic nerve rotation animals habitually moved the ipsilateral ear during blinking responses across the experiment (1 year), without any sign of adaptation. The EMG activity of the OO muscle in response to air puff presentations for the two types of anastomosis carried out here started 6–7 weeks after surgery, i.e. coinciding with the recovery of background activity in the EMG recordings.
Eyelid responses evoked by air puff stimulation of the cornea ipsilateral to the anastomosed side
Figure 2 illustrates some examples of blinks evoked in anastomosed and control animals by a puff of air presented to the ipsilateral (left, operated side) cornea and tarsal skin. Illustrated records were taken 3 and 9 months following anastomosis. The activation of the ipsilateral OO muscle in controls started 6–8 ms following air puff presentation, and was followed 4–5 ms later by a fast (up to 1500°/s), downward (20–25°) lid displacement. Thus, onset of lid downward displacement was 9–13 ms (10.6 ± 0.5 ms) after the onset of the air puff. The total duration of this downward movement of the lid was 20–25 ms for ipsilateral corneal stimuli. The initial downward movement of the lid was followed in controls by a succession of small downward sags (50 ms in duration). The blink ended by a slow upward movement that brought the lid back to an open position.
|
Three months after anastomosis, the downward lid movement evoked by a puff of air presented to the cornea of the operated side in animals with zygomatic nerve rotation (Fig. 2A, middle set of records) was
1.5 times larger than that evoked in controls (Fig. 2A, top). The same ratio in blink amplitude between the zygomatic nerve rotation animals and controls was still evident 9 months following surgery (Fig. 2B, middle versus top sets of records). The latency of the blink response recorded on the anastomosed side was always larger (2–3 ms) than that of controls, and never reached control values during the recording sessions (Fig. 3B). When compared with controls, the OO muscle of the zygomatic nerve rotation operated side presented a long-lasting, although disorganized, EMG response to air puff stimuli (Fig. 2A, middle), a fact still more noticeable 9 months after zygomatic nerve rotation and resuture (Fig. 2B, middle).
Results following buccal–zygomatic anastomosis were quite different from those of zygomatic nerve rotation and resuture. Three months after buccal–zygomatic anastomosis, the downward lid movement evoked by a puff of air was only about one-quarter of that evoked in control animals (Fig. 2A, bottom set of records). Moreover, this eyelid response could not be evoked by the EMG activity of the OO muscle, which was almost non-existent at that epoch. Nine months after buccal–zygomatic anastomosis, blinks evoked by ipsilateral air puffs were even larger, but, again, almost imperceptible blink-related EMG activity was recorded (Fig. 2B, bottom). Obviously, those blinks were evoked by a progressive increase in the activity of the eye retractor bulbi motor system (Delgado-García et al., 1990).
Figure 3 represents EMG and eyelid measurements carried out during ipsilateral air puff presentations throughout the 12 months following zygomatic nerve rotation and buccal–zygomatic anastomosis. No response could be recorded on the anastomosed side at month 0 (i.e. 5 days after surgery) because, as indicated, reinnervation of the anastomosed (left) OO muscle took place 6–7 weeks following the two types of anastomosis. For all recording sessions, onset latency of EMG activity in the OO muscle (Fig. 3A) and of lid downward movement (Fig. 3B) was significantly (P < 0.01 at least) shorter when air puff stimulation and recording were in controls than when they were in the anastomosed animals. No statistically significant trend in session evolution was found for the onset latency of EMG activity of the OO muscle or of eyelid movement in response to air puff stimulations of the cornea ipsilateral to the anastomosed side.
As already proved for the eyelid motor system (Evinger et al., 1991; Gruart et al., 1995a), the integrated EMG activity of the OO muscle is related linearly to maximum lid downward movement in response to air puff presentations, as can be seen comparing session evolution (Fig. 3C and D), in particular for controls and zygomatic nerve rotation animals. Thus, the repeated-measurement ANOVA showed that both the integrated EMG activity (Fig. 3C) and lid downward movement (Fig. 3D) had a similar and statistically significant (P < 0.0001) ascendant linear trend throughout the 12 month period of recordings in zygomatic nerve rotation animals. This progressive increase in the amplitude of reflex blinks evoked by the same stimulus presented to the ipsilateral cornea suggested the development of a compensatory hyper-reflexia. For 6–12 month recording sessions, both the integrated EMG activity of the OO muscle (Fig. 3C) and the maximum lid downward movement (Fig. 3D) evoked by ipsilateral air puffs in zygomatic nerve rotation animals were significantly (P < 0.01) larger than the corresponding values obtained from controls. In contrast, buccal–zygomatic anastomosed animals showed almost no noticeable integrated EMG activity for the duration of the experiment, but an increasing amplitude in eyelid downward displacement in response to ipsilateral air puff presentations was noticed (Fig. 3C and D). Because of the peculiar eyelid profiles (Fig. 2A and B, bottom), long time-to-peak response (Fig. 4) and absence of EMG activity in the OO muscle (Figs 2A and B, bottom, and 3C), these air puff-evoked blinks in buccal–zygomatic anastomosed animals were ascribed to a compensatory increase in the activity of the synergistic ipsilateral eye retractor bulbi motor system.
|
The fact that peak EMG activity of the OO muscle is related linearly to peak lid velocity during air puff-evoked blinks (Evinger et al., 1991
; Gruart et al., 1995a) was confirmed by the data illustrated in Fig. 3E and F, since the two parameters followed similar and statistically significant (P < 0.0001) ascendant linear trends through the recording sessions for controls and zygomatic nerve rotation animals. On the other hand, the peak EMG activity of the OO muscle (Fig. 3E) and the peak lid velocity of evoked blinks (Fig. 3F) recorded from controls were significantly (P < 0.05 at least) larger than the corresponding values recorded on zygomatic nerve rotation and buccal–zygomatic anastomosed animals. These results suggest that the increased air puff-evoked eyelid response observed following zygomatic nerve rotation was due to a long-lasting muscle activity (without any increase in the simultaneous firing of the innervating OO motoneurons, which would increase peak EMG values) and, probably, to the compensatory contribution of the retractor bulbi system (see double-headed arrows in Fig. 2B, middle). As indicated, eyelid responses in buccal–zygomatic anastomosed animals were entirely the result of eye retractor bulbi motor system activity.
Ipsilateral eyelid responses evoked by air puff stimulation of the peribuccal area ipsilateral to the anastomosed side
As illustrated in Figs 5 and 6, eyelid responses evoked in control animals by air puff stimuli presented to the peribuccal area were of larger latency (36.5 ± 5.4 ms versus 10.6 ± 0.5 ms, P < 0.001), and smaller amplitude (8.4 ± 1.6° versus 23.7 ± 3.2°, P < 0.001; see Fig. 6B) and peak velocity (215.0 ± 15.6°/s versus 1111 ± 82.6°/s) than those evoked by air puffs to the cornea. Both the integrated (15.3 ± 2.6 µV x s versus 27.9 ± 2.9 µV x s, P < 0.01; Fig. 6A) and peak (1.54 ± 0.8 mV versus 2.1 ± 0.3 mV, P < 0.05) EMG activity were smaller when the air puff was presented to the peribuccal area instead of cornea and periorbital skin.
|
Surprisingly, air puffs to the peribuccal area produced eyelid reflex responses with larger amplitude (28.6 ± 2.7° versus 8.4 ± 1.6°, P < 0.001; Figs 5, middle, and 6B) and greater peak velocity (317.0 ± 18.4°/s versus 215.0 ± 15.6°/s, P < 0.05) in cats with zygomatic nerve rotation than those evoked in controls. Also, the integrated (45.0 ± 2.8 µV x s versus 8.1 ± 1.6 µV x s, P < 0.01; Fig. 6A) EMG activity was higher in zygomatic nerve rotation animals than in controls. These results reinforce the earlier proposal (Gruart et al., 1996
) that axotomy of the zygomatic branch of the facial nerve produces a noticeable hyper-reflexia manifested here at the side ipsilateral to the lesion. Moreover, buccal–zygomatic anastomosed animals presented eyelid responses to peribuccal air puffs of similar amplitude to those evoked in controls (Fig. 6B). Since those eyelid responses were not evoked by the EMG activity of the reinnervated OO muscle (Fig. 6A), we have to assume that, in this case, eyelid displacement was the indirect effect of eye retraction produced by the retractor bulbi motor system, a fact also suggested by the long latency of these reflex responses (Fig. 5B, bottom).
Eyelid responses recorded on the side contralateral to the anastomosis
In control animals, the eyelid responses evoked on the right side by the same air puffs (3 kg/cm2, 100 ms) were equal to those evoked on the left side (Figs 2A, top, and 7A, top). In zygomatic nerve rotation animals, the eyelid responses evoked on the right (unoperated) side (i.e. the side contralateral to the anastomosis) by air puffs presented to the right cornea were significantly (P < 0.05 at least) larger in amplitude and duration than those recorded in controls (Figs 7A, middle, and 8A). In the same way, blinks recorded in buccal–zygomatic anastomosed animals on the side contralateral to the anastomosis, and evoked by air puffs presented to the same side, were significantly (P < 0.05 at least when compared 12 months after anastomosis) larger in amplitude and, mainly, in duration (measured as total lid displacement area, i.e. ° x s) than those evoked in controls (Figs. 7A, bottom and 8A).
|
Blinks evoked on the right (non-anastomosed) side and evoked by air puffs presented to the left (anastomosed) side were also recorded. In this case, the eyelid responses recorded in the two types of anastomosed animal were significantly (P < 0.01 at least) larger in amplitude and duration than those recorded in control animals (see Figs 7B and 8B). Finally, blinks evoked on the right (non-anastomosed) side by air puffs presented to the right peribuccal area yielded similar results, i.e. blinks on the non-operated side of all operated cats were significantly (P < 0.01 at least) larger in amplitude and duration than those evoked in controls (Figs 7C and 8C). Taken together, these data confirm the presence of hyper-reflexive blink responses on the side contralateral to the axotomy, in both types of anastomosis. Hyper-reflexive responses were observed to air puffs presented to sites ipsilateral and contralateral to the recording side. Since these responses increased mostly in duration, we have to assume that this was due to an increase in the duration of the Rap2 component of the air puff-evoked reflexes (see Gruart et al., 1995
a) and/or to the compensatory activation of the eye retractor bulbi motor system (Delgado-García et al., 1990).
Eye blink responses evoked by flash stimulation
Representative examples of eyelid responses evoked in both anastomosed (left) and non-anastomosed (right) sides by flashes of light are illustrated in Fig. 9. Light flashes evoked a brief EMG activation of the OO muscle with a latency of 34–36 ms in controls (Fig. 9A, top set of records). Lid downward movement started 4 ms after OO muscle activation. On the whole, light-evoked blinks presented a significantly (P < 0.0001) longer latency and smaller amplitude and peak velocity than those evoked by air puffs. Values obtained in the zygomatic nerve rotation animals were similar in latency to those of controls (Fig. 9A, middle). However, the percentage of responses (Fig. 9B), integrated (Fig. 9C) and peak EMG activity of the OO muscle, lid downward movement (Fig. 9D), and peak lid velocity on the anastomosed side during light-evoked blinks were significantly (P < 0.001 at least) smaller than the corresponding values for controls (Fig. 9B–D) throughout the experiment. As quantified in Fig. 9B–D, flash-evoked blinks recorded on the non-operated side in zygomatic nerve rotation animals were as in controls, with no sign of hyper-reflexia.
Flash-evoked blinks on the operated side of buccal–zygomatic anastomosed animals were (as described for air puff-evoked blinks) the result of activation of the eye retractor bulbi motor system. This is a true compensatory response, since accessory abducens motoneurons innervating the retractor bulbi system are not able to produce action potentials in control animals with a normal eyelid motor circuitry (Trigo et al., 1999b).
Eye blink responses evoked by tones
Control animals showed a very weak and variable response to tone presentation (not illustrated). The latency of the response was 45–50 ms, and the small (2–4°) evoked blink habituated following repeated (5–10 times) presentation of the same stimulus. Tone-evoked blinks presented a significantly (P < 0.0001) longer latency and smaller amplitude and peak velocity than those evoked by puffs of air. Following the two types of anastomosis carried out here on the (left) facial nerve, no tone-evoked response was recorded on the anastomosed side, although small blinks were observed on the non-operated side in response to tone presentation.
Spontaneous eyelid responses
Cats usually make a peculiar winking movement called ‘friendly display’ as a social communication signal (for description and references see Gruart et al., 1995a). An example of such bilateral eyelid response is shown in Fig. 0 (top set of records). These eyelid responses were noticeably distorted in profiles when performed by zygomatic nerve rotation animals, mainly on the operated side (Fig. 0, middle set of records). In those cases, the eyelid presented a wavy appearance instead of the smooth downward eyelid displacement observed in controls. Animals with buccal–zygomatic anastomosis presented an almost identifiable friendly response on the anastomosed side, that was probably the result of levator palpebrae muscle relaxation (arrow) during the lid response (Fig. 0, bottom set of records).
|
Lid movements during mouth-related motor activities
Lid movements on both control (A) and anastomosed (B) sides during the licking of a few milk drops applied directly to the tongue tip are illustrated in Fig. 1, for a buccal–zygomatic anastomosed animal. Licking produced a noticeable EMG activation of the OO muscle on the anastomosed (left) side, and no EMG activity in the OO muscle of the control (right) side. As a consequence, there was also an oscillation of the upper eyelid on the anastomosed side at a dominant frequency of 3–6 Hz (P < 0.001). No significant modification of this licking-induced oscillation of the lid position seemed to occur, since a similar pattern of response was recorded in the three buccal–zygomatic anastomosed animals up to the end of this study (1 year after the anastomosis). Also, Fig. 1 illustrates the asynchrony between the zygomatic and buccal branches of the facial nerve. Thus, when the animal made two spontaneous and successive blinks (downward arrows in Fig. 1A and B), the EMG activity of the OO muscle disappeared on the operated side, indicating an inhibition of buccal motoneurons during blink performance.
|
Frequency-domain properties of eye blinks
Figure 2 illustrates the frequency-domain characteristics of spontaneous and reflex eyelid responses recorded in both control and anastomosed animals. The power spectra of selected acceleration segments corresponding to eyelid position traces recorded during control reflex blinks showed a significant (P < 0.01) peak at 18–22 Hz (Fig. 2A, continuous curve in histogram). In contrast, the power spectra of similar reflex eyelid motor responses obtained from the two types of anastomosed animal showed a non-significant multipeaked profile. Indeed, the power at the dominant frequency for air puff-evoked blinks in controls was three orders of magnitude larger than that for blinks induced in the two types of anastomosed animal (Fig. 2A, dashed and dotted curves in histogram). Results were similar for air puffs presented to the peribuccal area ipsilateral to the anastomosed side (Fig. 9B). These data imply that reinnervation of the OO muscle following the two types of anastomosis used here disturbed eyelid performance, by the modification of its normally tuned resonant properties (Domingo et al., 1997
; Trigo et al., 1999b). The data also imply that a higher motor effort is probably necessary to move the reinnervated lid, a further reason to explain the observed hyper-reflexia.
Moreover, the fast Fourier transform analysis of 1.024 s acceleration recordings from the anastomosed (left) eyelid, taken while the buccal–zygomatic anastomosed animals licked a few drops of milk (see Fig. 1), showed a significant (P < 0.01) peak at 4–6 Hz. During licking, the power spectra of both controls and zygomatic nerve rotation animals showed no definite or significant peaks at any frequency (Fig. 2C). In this case, the spectral analysis of acceleration recordings during lid responses suggested that lid biomechanics were under the neural oscillatory control characteristic of tongue and buccal movements (4–6 Hz; see Gruart et al., 1996; Domingo et al., 1997), far below the typical frequency of normal lid spontaneous, reflex and learned responses (20 Hz; see Domingo et al., 1997; Trigo et al., 1999b).
Histological analysis
The location and distribution pattern of labelled motoneurons following HRP injection in the OO muscles of both control and anastomosed animals were checked. As already described (Shaw and Baker, 1983), the facial motoneurons innervating the OO muscle in the three control animals were packed together in the dorsal subdivision of the facial nucleus (Fig. 3, top). The mean number of labelled motoneurons from the three muscles injected was 884 ± 27. These values are similar to those recently reported for cat OO motoneurons labelled with the same HRP procedure in a different experiment (Gruart et al., 1996). Labelled OO motoneurons had the characteristic multipolar shape, with 4–7 principal dendrites.
|
The number of labelled motoneurons following HRP injection in the OO muscle of the zygomatic nerve rotation anastomosed side (915 ± 35) was slightly, but not significantly, larger than control values (Fig. 3, middle). Interestingly, no labelled motoneuron was observed outside the dorsal subdivision of the facial nucleus, indicating that nearby motoneurons did not invade the OO muscle after zygomatic nerve rotation anastomosis, and that the observed functional changes originated from the original pool of motoneurons. Those motoneurons labelled in zygomatic nerve rotation animals showed an appearance similar to that of controls.
In the case of buccal–zygomatic anastomosed animals, labelled motoneurons appeared concentrated at the lateral and ventrolateral subdivisions of the facial nucleus, with a total number of 789 ± 45 cells (Fig. 3, bottom). Those motoneurons showed multipolar shapes, with 4–8 principal dendrites. No labelled motoneuron was observed inside the dorsal subdivision of the facial nucleus, indicating that no OO motoneuron was able to reinnervate the OO muscle following its implantation into ear muscles.
In contrast to previous descriptions following hypoglossal–facial anastomosis (Angelov et al., 1993, 1997) or re-anastomosis of the main facial nerve trunk in rats (Angelov et al., 1996; Streppel et al., 1998), hyperinnervation of the OO muscle by facial motoneurons was observed in the present experiments on cats.
|
Discussion |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Reflexively evoked eye blink responses
Air puffs, flashes of light and tones are used commonly to evoke reflex blinks in mammals (including humans) for functional studies of the mimic musculature, and for the classical conditioning of the nictitating membrane/eyelid response (Marquis and Porter, 1939
; Gormezano et al., 1983; Evinger et al., 1991; Gruart et al., 1995a; Kim and Thompson, 1997). The latency, amplitude, profile and frequency-domain properties of control values reported here for air puff-, light- and tone-evoked reflex blinks are in the range of those found in previous reports in cats (Gruart et al., 1995a, 1996; Domingo et al., 1997). According to the present results, reflex blinks do not recover control values following buccal–zygomatic anastomosis of the (left) facial nerve, at least within the period of this study (1 year). The results also suggest that trigeminal, visual and acoustic signals arriving at the different subdivisions of the facial nucleus are unable to generate a reorganization of their presynaptic terminals to redistribute the arriving sensory signals to the new disposi tion of facial nucleus motor pools following buccal–zygomatic anastomosis. In contrast, zygomatic nerve rotation anastomosis allowed the performance of reflex blinks and spontaneous eyelid responses, but with noticeable alterations in eyelid profiles and kinematics, and accompanied by a significant bilateral hyper-reflexia. This hyper-reflexia was apparently restricted to trigeminal pathways, because it was not evoked by flashes or tones.
Although crossed facial–facial anastomosis is not usually carried out in humans, some comparison could be made with available results regarding hypoglossal–facial anastomosis, which today is a standard method for the treatment of facial paralysis after irreversible damage of the cranial portion of the facial nerve (Körte, 1903; Stennert, 1979; Miehlke et al., 1981; Hammerschlag, 1987; May et al., 1991), and has been used as a model for studying neuronal plasticity in animal experimentation (Neiss et al., 1992; Gruart et al., 1995b, 1996). The buccal branch of the facial nerve has to share many neural commands with the hypoglossal nerve in order to achieve a coordinated displacement of the mouth cavity during phonation, mastication and deglutition (Dinardo and Travers, 1994; Gruart et al., 1996). Results related to the evolution of reflex blinks following ipsilateral hypoglossal–facial anastomosis in man are quite different. For example, some authors reported the recovery of the early (R1, following Kugelberg, 1952) component of the blink evoked by electrical stimulation of the ipsilateral supraorbital nerve (Kilimov and Linke, 1978; Willer et al., 2002), while others observed a recovery exclusively in the late (R2, following Kugelberg, 1952) components of the reflex eyelid response (Vera et al., 1975), or even some plastic changes at the trigeminal nucleus (Tankéré et al., 2000). Finally, the recovery of a normal blink reflex following hypoglossal–facial anastomosis in humans frequently has been ascribed to the reinnervation of the OO muscle by axons originating from the facial nerve (Struppler and Dobbelstein, 1963; Iansek et al., 1986), to the presence of axon reflexes simulating blink (R1) responses (Montero et al., 1996) or even to pre-existing trigeminal inputs on hypoglossal motoneurons (Stennert and Limberg, 1982). In a recent long-term study (7 months) in cats, it was reported that reflex blinks are not recovered following a successful hypoglossal–facial anastomosis (Gruart et al., 1996). As already indicated by Sperry (1945, 1947) in a broader statement, the present results and the available information from clinical studies suggest that both hypoglossal and buccal motor neural circuits are unable to elaborate a proper reflex blink when the corresponding (hypoglossal and buccal) motoneurons are forced to innervate the OO muscle in adult mammals.
The evolution of reflex blink responses following facial–facial anastomosis indicated the appearance of a hyper-reflexia in response to air puff presentations, extensible to both the control and the anastomosed side. Similar phenomena, including hyper-reflexia, hemifacial spasms, synkinesis and increased axonal reflexes, have also been reported in human patients with facial nerve injuries (Montero et al., 1996; Spector 1997). It has been proposed (Gruart et al., 1996) that the experimental paralysis of the eyelids modifies the activation threshold of corneal receptors on the anastomosed side. The increase in activity of corneal receptors could contribute to the bilateral hyper-reflexia reported here in response to air puffs. Data suggesting some plasticity in the corneal nociceptive pathway have been reported in rats (Pozo and Cerveró, 1993), which could explain why no hyper-reflexia was observed for light- and tone-evoked reflex blinks (see also Gruart et al., 1996). In addition, the increased blink responses to air puffs applied to the lower face after zygomatic nerve axotomy suggest the presence of some facilitatory common relay neurons located in the trigeminal nucleus. For example, it has been described that corneal anaesthesia removes tonic facilitatory inputs on blink-related second order trigeminal inputs, making them functionally insensitive to input signals from exteroceptors located in the chin (Trigo et al., 1999a).
As already reported in cats following a hypoglossal–facial anastomosis, the small recovery of reflex blink responses could be ascribed either to a modest increase in the activity of reinnervating buccal motoneurons or to an increase in the gain of trigeminal pathways impinging upon accessory abducens motoneurons and extraocular motoneurons (Gruart et al., 1996). The latter two motor systems are also able to produce a downward movement of the upper lid by retracting the eyeball back into the orbit (Delgado-García et al., 1990).
In this regard, the presence of late downward eyelid movements in both types of anastomosis, as well as its characteristic time course, suggests the involvement of the cat’s retractor bulbi motor system to increase eyelid downward displacement. This phenomenon represents a compensatory response (see Delgado-García, 1998) involving a parallel motor system (not affected by the surgery), usually inactive for the range of air puff presentations used here. Similar results were obtained during hypoglossal–facial anastomosis (Delgado-García et al., 1990; Gruart et al., 1996).
It is possible to suggest that the hyper-reflexia observed in the present experiments could also be related to the episodes of hemifacial spasm described in humans following facial nerve section and resuture (Kuroki et al., 1994). Indeed, EMG recordings of OO muscle in anastomosed cats showed frequent episodes of tonic basal activity, that could be involved in both hyper-reflexia and muscle spasms.
Adaptability of post-embryonic facial motor circuits to new motor tasks
The OO muscle is a flat, highly organized and regionalized muscle (Gordon, 1951; McLoon and Wirtschafter, 1991). Although facial motoneurons innervating the OO muscle are concentrated in the dorsal part of the facial nucleus, recent studies suggest a further organization of this motor pool, by which facial motoneurons innervating the pre-tarsal, pre-septal and pre-orbital parts of the OO muscle are separated in the dorsal subnucleus (Trigo et al., 1999b). These morphological data suggest a further specification of synaptic inputs to each subdivision of OO motor units, probably related to specific motor functions. For example, it has been shown that motor units located in the pre-tarsal region present a phasic firing and are involved preferentially in reflex blink responses, while those occupying the pre-septal and pre-orbital regions are able to display some tonic firing and are involved preferentially in motor responses involving the facial expression of emotions (Gordon, 1951; Trigo et al., 1999b). This specificity in motoneuronal projections and the restricted location of particular OO motor units explain why even the simple 180° rotation and resuture of the zygomatic nerve is able to disturb the fine performance of spontaneous and reflexively evoked eyelid responses.
The final molecular and subcellular steps following the acquisition of new motor skills mean the reorganization of pre- and postsynaptic specializations (see Bliss and Collingridge, 1993; Edwards, 1995; Dobkin, 1997; Schwab, 2002). Apparently, these plastic processes are not taking place at motoneuronal level, nor are they restricted to the local microenvironment of the specific pool of motoneurons involved in each precise facial movement.
The disturbance of the specifically tuned frequency-domain properties of OO motor units (Trigo et al., 1999b) by the two types of anastomosis used here produced unwanted oscillations of the upper eyelid during spontaneous and reflex blinks (see Fig. 0
), and abnormal lid responses during licking and eating (see Figs 1 and 12). As already described (Domingo et al., 1997), buccal and tongue movements in cats during drinking and eating oscillate at frequencies lower (4–7 Hz) than those characterizing lid responses (20 Hz), a fact easily noticed during licking in animals with a buccal–zygomatic anastomosis.
The present results give partial support to the proposed theory that the OO muscle hyperinnervation of a dual origin (facial and hypoglossal) observed following hypoglossal–facial anastomosis could be induced by mismatching between the motor function of the OO muscle and the discharge properties of reinnervating motor cells (Angelov et al., 1993). Such a mismatch would force the improperly driven muscles to continue the secretion of putative reinnervation-promoting factors (Angelov et al., 1993; Evans, 2000). Indeed, the reported hyperinnervation of the OO muscle following hypoglossal–facial anastomosis (Gruart et al., 1996) was not observed in the present experiments. Nevertheless, the amount in the reinnervation pattern cannot be ascribed to the appropriateness of neural control signals arriving at muscle fibres, but, more probably, to the presence of promoting factors determined at early stages in the development of brainstem motor systems.
Conclusions
The present results further support the contention that brainstem motor systems of adult mammals are organized in a fixed manner (Sperry, 1945, 1947; Baker, 1985; Gruart et al., 1995b, 1996) and cannot adapt to motor tasks of different kinematics and timing. The supposed ‘adaptive plasticity’ of the physiological responses of axotomized motoneurons when reinnervating a foreign muscle (Gruart et al., 1996; and present experiments) or the ‘vestibular compensation’ following unilateral labyrinthectomy (Dieringer, 1995; Delgado-García, 1998) have to be ascribed to compensatory phenomena involving parallel, non-directly affected pathways (Baker, 1985), or to the participation of higher neural centres able to restore motor disarrangements through re-educative learning (Sperry, 1945, 1947). Recent studies on motor cortex reorganization following facial nerve damage support the latter contention (Franchi, 2000a, b).
|
Acknowledgements |
|---|
We wish to thank Mr Roger Churchill for help in the editing of the manuscript. This work was made possible by cooperation grants from Acciones Integradas Hispano-Alemanas (DAAD/MEC, HA1996-0128), EU Cost B-10 ‘Brain Damage and Repair’, and grants from the Spanish DGICYT (PM98-011), Fundació La Caixa (00/032-00) and Junta de Andalucía (CVI-122) in Spain, and from the Jean Uhrmacher Foundation in Germany.
|
References |
|---|
|
TopSummaryIntroductionMaterial and methodsResults |
|---|