Region Diminishes Polynnervation of the Motor
Endplates and Improves Recovery of Function After
Hypoglossal Nerve Injury in Rats
Emilia Evgenieva, PhD, Patrick Schweigert, Orlando Guntinas-Lichius, MD, PhD,
Stoyan Pavlov, MD, Maria Grosheva, MD, Srebrina Angelova, MD, Michael
Streppel, MD, PhD, Andrey Irintchev, MD, PhD, Emmanouil Skouras, MD,
Stefanie Kuerten, MD, Nektarios Sinis, MD, Sarah Dunlop, PhD,
Victoria Radeva, MD, PhD, and Doychin N. Angelov, MD, PhD
Background. Using the rat facial nerve axotomy model, the
authors recently showed that manual stimulation of denervated whiskerpad muscles reduced the posttransectional polyinnerva-tion at the neuromuscular juncpolyinnerva-tions and promoted full recovery of vibrissal whisking. Objective. Prompted by implications for rehabilitation therapy, the authors examined whether manual stimulation of denervated supra- and infrahyoid muscles would also improve recovery after unilateral lesion on the hypoglossal nerve. Methods. Adult rats underwent transection of the right hypoglossal nerve. Half of the animals received no postoperative treatment, and the other half were subjected to daily manual stimulation of the suprahyoid/sublingual region for 2 months. Recovery was assessed by measuring the angle of tongue-tip devi-ation from the midline, degree of collateral axonal branching at the lesion site (counts after retrograde labeling with 2 fluorescent dyes), synaptic input to the hypoglossal motoneurons using synap-tophysin immunocytochemistry, tongue-muscles motor represen-tation in the cerebral cortex after c-Fos immunocytochemistry, and
portion of polyinnervated neuromuscular junctions. Results. In animals receiving manual stimulation, the tongue-tip deviation was 37.0 ± 49.37°, whereas values in control nonstimulated rats were significantly higher (50.1 ± 9.01°; P < .05; mean ± SD). Improved recovery was not associated with reduced collateral axonal branching; there were also no differences in tongue-mus-cles representation in the motor cortex. However, manual stimu-lation restored the total synaptic input to levels in intact animals and reduced the proportion of polyinnervated neuromuscular junctions compared with nonstimulated animals. Conclusion. The data show that manual stimulation of denervated muscles improves functional outcome following peripheral nerve injury. This suggests immediate potential for enhancing clinical rehabilitation strategies.
Key Words: Motoneuron—Axotomy—Hypoglossal nerve—Axonal sprouting—Motor cortex—Suprahyoid-sublingual region— Neuro-muscular junction—Polyinnervation
D
espite considerable clinical advances in micro-surgery, functional recovery in humans after transection and suture of peripheral nerves remains poor, and the occurrence of a “postparalytic syndrome” (paresis, synkinesis, and dysreflexia) is inevitable.1Insufficient recovery has been mostlyattrib-uted to the misrouting of regrowing axons that fail to rejoin their original nerve fascicles2 and produce
numerous axonal sprouts both at the transection site (“collateral axonal branching”3) and within target
mus-cles (“intramuscular or terminal axonal sprouting”4).
We have previously characterized events that influ-ence functional recovery following facial nerve transec-tion and suture in the rat. Quantitative assessment of vibrissal motor performance followed by triple retro-grade labeling of facial motoneurons in the same ani-mals5indicated that the main factor limiting restoration
of functional performance was a high degree of intra-muscular sprouting, which was associated with a high proportion of polyinnervated motor endplates.6,7
From the Faculty of Pre-School and Primary School Education, Sofia University, Bulgaria (EE, VR); Department of Anatomy I University of Cologne, Germany (PS, SK, DNA); Department of Oto-Rhino-Laryngology, University of Jena, Germany (OG-L); Department of Anatomy, Histology, Embryology, Medical University Varna, Bulgaria (SP); Department of Oto-Rhino-Laryngology, University of Cologne, Germany (MG, SA, MS); Centre for Molecular Neurobiology, University of Hamburg, Germany (AI); Department of Trauma and Reconstructive Surgery, University of Cologne, Germany (ES); Department of Hand-, Plastic-, and Reconstructive Surgery with Burn Unit, BG-Trauma Centre, University of Tuebingen, Tuebingen, Germany (NS); and School of Animal Biology and Western Australian Institute for Medical Research, The University of Western Australia (SD). EE and PS con-tributed equally and share first authorship. AI and DNA concon-tributed equally and share last authorship.
Address correspondence to Doychin N. Angelov, MD, PhD, Institut I für Anatomie der Universität zu Köln, Joseph-Stelzmann-Strasse 9, D-50924 Köln, Germany. E-mail: angelov.anatomie@uni-koeln.de. Evgenieva E, Schweigert P, Guntinas-Lichius O, et al. Manual stimulation of the suprahyoid-sublingual region diminishes polynnervation of the motor endplates and improves recovery of function after hypoglossal nerve injury in rats. Neurorehabil Neural Repair. 2008;22:754-768. DOI: 10.1177/1545968308316387
Studies in experimental animals have shown that mild electrical stimulation of the denervated soleus muscle inhibits intramuscular sprouting and diminishes motor endplate polyinnervation.8In addition, soft tissue
mas-sage has been shown clinically to have several benefits.9
The findings prompted us to examine the effect of man-ual stimulation (MS) on both functional recovery of vibrissal muscles and the degree of polyinnervation fol-lowing facial nerve injury. Video-based motion analysis of vibrissal motor performance showed that daily MS for 2 months resulted in full recovery of whisking, which was associated with reduced polyneuronal rein-nervation of motor endplates and not with attenuation of misdirected axonal regrowth.10
Because MS of vibrissal muscles significantly improved functional recovery of whisking after injury to the facial nerve, we asked whether this rehabilitation approach would be also successful after injury to another nerve— namely, the hypoglossal nerve. An additional impetus for this study was the recent finding that motor control of human tongue movements can be improved by selected tongue training techniques.11
MATERIALS AND METHODS
Materials
Fluoro-Gold (FG) was purchased from Fluorochrome, Inc (Denver, Colorado), Fast Blue (FB) from EMS-Chemie GmbH (Groß-Umstadt, Germany), and Alexa Fluor 488-conjugated α-bungarotoxin from Molecular Probes (Leiden, The Netherlands). Rabbit polyclonal anti-synaptophysin was purchased from Biometra (Göttingen, Germany), and polyclonal antineuronal class III β-tubu-lin antibody was from Covance (Richmond, California).
Animals, Groups, and Overview of
Experiments
Seventy-eight young female adult (175-200 g) Wistar rats (strain HsdCpb:WU, Harlan-Winkelmann, Borchen, Germany) were divided into 2 control and 4 experimen-tal groups (Table 1). We used only female rats because testosterone has been shown to beneficially affect peripheral nerve regeneration.12
Before and after surgical treatment, rats were fed standard laboratory food (Ssniff, Soest, Germany) and provided tap water ad libitum and an artificial light-dark cycle of 12 hours light on, 12 hours off. All experi-ments were conducted in accordance with the German Law on the Protection of Animals, and procedures were approved by the local Animal Care Committee.
Groups 1 to 3 (n = 16 per group) were used to study collateral axonal branching at the site of lesion, synaptic input to the hypoglossal motoneurons, and the quality of target reinnervation (m. hyoglossus). Group 1 consisted of intact animals. All rats in groups 2 and 3 were subjected to unilateral transection and suture of the right hypoglos-sal nerve (hypogloshypoglos-sal-hypogloshypoglos-sal anastomosis [HHA]). Animals in group 2 received no postoperative treatment, whereas those in group 3 were subjected to MS of the extrinsic and intrinsic suprahyoid-sublingual region.
In addition to misdirected reinnervation of muscle targets, insufficient recovery has also been attributed to rearrangement of cortical representations.13,14 Cortical
tongue muscle representation volume was therefore examined (groups 4-6; n = 10 per group). Animals in group 4 (intact) were subjected to right unilateral HHA and were kept under anesthesia for 1 hour prior to per-fusion fixation. Animals from groups 5 and 6 under-went HHA and survived for 2 months; those in group 5 received no postoperative treatment, whereas those in Table 1. Experimental Design Chart Depicting Animal Grouping and Procedures (eg, Hypoglossal-Hypoglossal Anastomosis [HHA]), With or Without Manual Mechanical Stimulation of the Tongue Muscles (MS)
Degree of Collateral Reinnervation
Restoration of the Axonal Branching Extent of Synaptic Pattern of the Tongue Muscles Tongue Position as Estimated by Input to Motor Endplates Representation by Measuring the Double-Retrograde Hypoglossal in the Hyoglossus Volume
Group of Animals Deviation Angle Labeling Motoneurons Muscle (CTMRV)
Group 1: Intact 16 8 8 8
Group 2: HHA only 16 8 8 8
Group 3: HHA + MS 16 8 8 8
Group 4: Intact 10 10
Group 5: HHA only 10 10
Group 6: HHA + MS 10 10
All animals from groups 1 to 3 were subjected to postoperative measurement of the tongue-tip deviation from the midline. Thereafter, half were used for estimating the degree of collateral axonal branching and the other half for establishing the extent of synaptic input to the hypoglossal motoneurons and the pattern of the motor endplate reinnervation. The animals from groups 4 to 6 served to establish changes in cortical tongue muscles representation volume (CTMRV).
group 6 received MS exactly as those in group 3. After 2 months, the right hypoglossal nerve of all rats in groups 5 and 6 was transected proximally to the earlier lesion kept under anesthesia for 1 hour prior to perfusion fixation.
Estimation of the deviation angle of the tongue-tip from the midline, degree of axonal branching, synaptic input to the hypoglossal motoneurons, pattern of motor endplate reinnervation, and determination of cortical tongue muscle representation volume were undertaken at 2 months after surgery. Data for rats receiving MS were compared with those that did not receive MS.
All animals in groups 1 to 6 were used to determine the deviation angle of the tongue-tip from the midline, a stan-dard procedure for estimating hypoglossal nerve func-tion.15Thereafter, half the animals (n = 8) in groups 1 to 3
were used to establish the degree of collateral axonal branching by means of double-retrograde neuronal label-ing (see below). The remainlabel-ing rats in groups 1 to 3 (n = 8) were used to determine the synaptic input to the hypoglossal motoneurons (using immunocytochemistry for synaptophysin; see below) and the proportion of mono- and polyinnervated motor endplates in the ipsilat-eral hyoglossus muscle (using immunocytochemistry for neuronal class III β-tubulin and histochemistry with alpha-bungarotoxin; see below). All animals in groups 4 to 6 (group 4: intact; group 5: HHA, no MS; group 6: HHA + MS) were used to determine the deviation angle of the tongue-tip from the midline and the cortical tongue muscle representation volume (see below).
Surgery
Transection and end-to-end suture of the right hypoglossal nerve (HHA) was performed by Dr O. Guntinas-Lichius, Dr M. Streppel, and Dr M. Grosheva after an intraperitoneal injection of ketamin/xylazin (100 mg Ketanest, Parke-Davis/Pfizer, Karlsruhe, Germany, and 5 mg Rompun, Bayer, Leverkusen, Germany, per kg body weight). The right hypoglossal nerve was exposed and transected proximal to its bifurcation into lateral and medial branches (Figure 1a). End-to-end suture (HHA) was performed immediately using two 11-0 atraumatic sutures (Ethicon, Norderstedt, Germany) and the wound closed using three 4-0 skin sutures (Ethicon).
Manual Stimulation of the Extrinsic and
Intrinsic Suprahyoid-Sublingual Region
On the day following surgery, the suprahyoid-sublingual region of all 16 animals from groups 3 and 6 were manually
stimulated. Manual stimulation was performed by gently stroking the lower jaw and upper neck to stimulate all 3 extrinsic (skeletal) muscles of the tongue (m. styloglos-sus, m. genoioglosstyloglos-sus, m. hyoglossus) for 5 minutes a day, 5 days a week, for 2 months. The pattern of manual stimulation mimicked the natural active movements during swallowing (Figure 1b). Animals rapidly became accustomed to this procedure within 2 to 3 days and did not show any signs of stress such a freezing or trying to bite, weight loss, or lack of grooming; rather, animals readily cooperated.
In addition, upon return to the cage, each rat that had received MS also had a drop of honey placed on its back, which was accessible to its cage companions. By licking away the honey, cage companions stimu-lated their intrinsic suprahyoid-sublingual region (m. longitudinalis sup., m. longitudinalis inf. m. trans-verses, and m. verticalis) for a further 5 to 10 minutes after MS.
Figure 1. Surgical procedure and postoperative treatment. (a) Schematic drawing of the rat hypoglossal nerve. Arrow points at the site of transection and suture. Adapted from Greene EC. Anatomy of the rat. Trans Am Philos Soc. 1935;27:1-370. (b) Manual “submental” stimulation of the extrinsic suprahyoid-sublingual region.
Restoration of Tongue Position
During Protrusion as a Sign for
Recovery of Function
Recovery of tongue function was estimated by mea-suring deviation of the tongue-tip from the midline (ie, the angle between the long axis of the tongue and the median line of the body running between the incisor teeth). Animals were held gently by an experimentator, and the upper lip was slightly lifted. Photographs in the frontal plane of all rats from an identical distance (about 20 cm) were taken using the macro-menu of a Nikon 50D digital Camera. The very high resolution of the pictures allowed us to readily identify the tip of the tongue as well as the long axis of the organ. The identi-cal position of each animal when photographed and the short distance between the camera and the head of the rat reduced the possible parallax errors maximally.
In intact rats, tonus of the right and left protruders (the extrinsic m. genioglossus and the intrinsic, vertical, and transverse muscles) is identical, and therefore the tip of the tongue was situated exactly in the midline behind the lower incisors (ie, the deviation from the midline was 0 degrees; Figure 2a). Following right-sided HHA, mal-function of all right protruders resulted in domination of the opposite (left) muscles, which displaced the tongue-tip to the right16(ie, the long axis of the tongue
no longer coincided with the midline; Figure 2b).
Estimation of Axonal Branching by
Double-Retrograde Labeling
Previous data after immunostaining of 50-μm thick vibratome sections for neuron-specific enolase (NSE; ie, no retrograde labeling performed) showed that the intact hypoglossal nucleus contained 3576 ± 284 NSE-immunoreactive perikarya; there were no significant changes in these values either at 1 (4010 ± 245) or 8 weeks after HHA (3412 ± 348).17
Application of fluorescence tracers. Eight rats from groups
1 to 3 were used to establish the degree of collateral axonal branching at the lesion site (HHA). Under Rompun/Ketanest anesthesia, the right hypoglossal nerve was reexposed distally to the suture site. The medial and lateral branches were transected and instilled with crystals of the retrograde fluorescent dyes (FB and FG, respectively; Figure 3a). Crystals were left in situ for 30 minutes, after which the application sites were care-fully rinsed, dried, and the wound closed. Ten days later, animals were fixed by perfusion (4% paraformaldehyde in 0.1 M phosphate-buffered saline, pH 7.4), and the brainstems were sectioned coronally at 50 μm.
Under normal physiological conditions, the hypoglos-sal nerve controls tongue movements by means of its 2 functionally different nerve branches.15 The medial
branch contains the axons of neurons in the ventral hypoglossal subnucleus and innervates muscles that are related to protrusion of the tongue (the extrinsic genioglossus and the intrinsic vertical and transverse muscles). The smaller lateral branch contains the axons of perikarya in the dorsal hypoglossal subnucleus and innervates muscles related to tongue retraction (extrin-sic styloglossus and hyoglossus and intrin(extrin-sic superior and inferior longitudinal).
After transection of the hypoglossal nerve, the regrow-ing axons navigate poorly and fail to rejoin their original nerve branches (medial or lateral) and therefore their Figure 2. Measurement of tongue-tip deviation from the midline (ie, of the angle between the long axis of the organ and the median line of the body running between the incisor teeth) in representative animals. The edges of the tongue are outlined by a dotted line. (a) In intact rats, the identical tonus of the right and left protruders and transverse muscles situated the tip of the tongue exactly in the middle behind the lower incisors (ie, the deviation from the midline was 0 degrees). (b) In operated animals, the left protruder dominated and dis-placed the tongue-tip to the right (ie, the long axis of the organ was no more covering the median line, and the angle between them was proportional to the recovery of function).
correct muscle targets. The aim of this procedure was not to determine whether myotopic organization of the hypoglossal nucleus had been preserved or restored but rather to establish the degree of collateral axonal branching at the lesion site by means of double-retrograde labeling and neuronal counts (see below).
Fluorescence microscopy. Sections were observed using an
epifluorescence microscope (Axioplan, Zeiss, Oberkochen, Germany). To restrict the fluorescence shine-through between the tracers, we used a custom-made band pass-filter set combination (AHF Analysentechnik, Tübingen, Germany). Separate color images of retrogradely labeled hypoglossal motoneurons were captured through the
different filter sets using a CCD Video Camera System (SPOT, Visitron Systems, D-82178 Puchheim, Germany) and image analysis software (Optimas 6.5, Optimas Corporation, Bothell, Washington). “FG-masks” of all FG-labeled motoneurons were generated using Optimas: frames were binarized, dilated, and the outlines of each FG-labeled cell marked. FG-masks were then superim-posed over the FB-picture using the arithmetic options dia-logue from the image menu. The technique allowed us to readily identify cells stained by FGonly, FBonly, and all those double stained by FG + FB (Figure 4a,b), which were then counted manually on the computer screen.18 Although
time-consuming, the procedure was precise and allowed us to quantify the degree (index) of axonal branching. The index of axonal branching represents the ratio between motoneurons projecting branched axons into the medial and lateral branch to all motoneurons sending axons through both branches (ie, the percentage of double-labeled motoneurons). Rats with an intact hypoglossal nerve trunk subjected only to surgery for tracer application had an index of axonal branching of 0% (Figure 3b). Figure 3. Retrograde neuronal labeling in intact rats with 2
fluorescent dyes. (a) Schematic drawing of the rat hypoglossal nerve. The blue arrow points at the medial branch that was tran-sected and instilled with crystals of Fast Blue (FB) and the yellow
arrow at the lateral branch of the hypoglossal nerve (transected
and labeled with crystals of Fluoro-Gold [FG]). Adapted from Greene EC. Anatomy of the rat. Trans Am Philos Soc. 1935;27:1-370. (b) Myotopic organization of the hypoglossal nucleus in intact rats. Application of FB to the transected medial branch labeled perikarya, which were localized in the ventral hypoglos-sal subnucleus. Likewise application of FG to the transected lat-eral branch labeled perikarya, located in the dorsal hypoglossal subnucleus. No double-labeled perikarya were observed (ie, the degree of axonal branching was 0%).
Figure 4. Retrograde neuronal labeling after hypoglossal-hypoglossal anastomosis (HHA) in (a) nonstimulated and (b) stimulated rats. The organization of the hypoglossal motoneu-rons into subnuclei is evident no more, and due to collateral axonal branching, there appear double-labeled (Fast Blue [FB] + Fluoro-Gold [FG]) neuronal somata (arrows).
Counting. Counts of all hypoglossal perikarya labeled
with FB, FG (Figure 3b), and FB + FG (Figure 4a,b) were undertaken using the fractionator principle examining every fourth section through the hypoglossal nucleus. Details have been described previously.17Counting was
performed blindly with respect to treatment.
Measurement of Motoneuron Soma Sizes
Earlier work has shown that within 1 week after rein-jury of chronically axotomized mouse facial motoneu-rons, their atrophic cell bodies increase in size, and expression of growth-related proteins is enhanced.19
Thus, motoneuron size after a period of recovery from nerve transection and repair followed by a second axo-tomy is considered to reflect regenerative capacity and thus the functional state of regenerated motoneurons. Similarly, 3 months after femoral nerve transection and repair in mice, the degree of motor recovery correlates with soma size of regenerated motoneurons.20We
there-fore measured hypoglossal motoneuron area following retrograde labeling in rats with and without MS. Labeled hypoglossal motoneurons were photographed with a SPOT-CCD Video Camera System mounted on an Axioplan Zeiss microscope under ×16 magnification. Images were saved in an uncompressed format (TIFF). Analysis was performed with software ImageJ v. 1.38t (US National Institutes of Health, Bethesda, Maryland, http://rsb.info.nih.gov/ij/). An average of 60 motoneurons per hypoglossal nucleus was randomly selected and outlined semiautomaticaly using the Multi Cell Outliner plugin (Figure 5). The areas of each perykarion were automati-cally measured in μm2using the building analysis
func-tions of the software. Measurements were performed by 1 observer (S. Pavlov) who had no information about the postoperative treatment of the rats.
Establishment of the Synaptic Input to the
Hypoglossal Motoneurons
To compare the degree of synaptic input in rats with and without MS, we measured and established the over-all intensity of fluorescence in the hypoglossal nucleus after immunostaining for synaptophysin.
Tissue preparation. Perfusion fixed (4%
paraformalde-hyde) brainstems were cut coronally in 30-μm thick vibratome sections. Immunocytochemical staining for synaptophysin (rabbit polyclonal antisynaptophysin, Biometra, cat. no. 100-599) was performed on every
fifth section through the hypoglossal nucleus in 1 incu-bation batch for all 24 rats.
Immunocytochemistry. Sections were immunostained
on a shaker at room temperature using 5.0% (w/v) bovine serum albumin (BSA, Sigma no. A-9647) in Tris-buffered saline (TBS) for 30 minutes; 1:4000 antisynap-tophysin, diluted in TBS plus 0.8% (w/v) BSA for 2 hours; 5.0% (v/v) normal sheep serum (NSS, Sigma no. S3772) plus 0.8% BSA in TBS for 15 minutes; and incu-bation with antirabbit IgG Cy3 conjugate (1:400; Sigma no. C-2306) in TBS plus 0.8% NSS for 1 hour.
Overall amount of synaptic terminals in the hypoglossal nucleus. To quantify pixel brightness, images were captured
with a slow-scan CCD camera (SPOT RT, Diagnostic Instruments, Inc, Sterling Heights, Michigan;×16 objec-tive) using Image-Pro Plus Software (Version 5.0; Media Cybernetics, Inc, Silver Spring, Maryland; Figure 6a-d). Black levels were kept constant, but gain was manipulated for each group, thereby ensuring that only a few pixels were saturated at the 255-pixel gray value. Each pixel therefore contained 8 bits of information encoding brightness ranging in value from 0 to 255. The scale for pixel brightness, or pixel gray value, was constructed so that the higher numbers indicate greater pixel brightness. The use of the collection filter further reduced the number of pixels saturated at 255. Thus, the background intensities were identical from image to image around a pixel gray value of 50. Accordingly, the level for measuring pixel number and brightness was set at 51 (Figure 6e). Figure 5. Motoneuron soma size. (a) Normalized frequency distributions of soma areas of back-labeled motoneurons in rats subjected to retrograde labeling only (no HHA) or to retrograde labeling 2 months after nerve repair and no stimulation (HHA) or nerve repair and manual stimulation (HHA + MS). The dis-tribution in the HHA + MS group differs from those in the other 2 groups (P < .001, Kolmogorov-Smirnov test). Number of motoneurons/rats studied is indicated in the panel. (b) Group mean values + SEM of soma areas of the motoneurons shown in A. Asterisk indicates significant difference from both other groups (P< .05, analysis of variance with Tukey’s post hoc test).
Analysis of Target Muscle Reinnervation
The ratio mono- versus polyinnervated motor end-plates was performed as described previously.7 We
selected the hyoglossus muscle rather than the m. genioglossus and m. styloglossus. The hyoglossus mus-cle extends as a thin musmus-cle sheet from the hyoid bone and enters the tongue laterally, between the masseter
and stylohyoideus muscles, allowing its easy identifica-tion and dissecidentifica-tion (Figure 7a,b).
The hyoglossus muscles were dissected free, cryopro-tected in sucrose, and cut longitudinally (30 μm) on a cryostat. Axons were immunostained with a rabbit poly-clonal antibody against neuronal class III tubulin (Covance, no. PRB-435P, 1:1000) and Cy3-conjugated antirabbit IgG (1:400; Sigma, Deisenhofen, Germany); Figure 6. Quantification of synaptic terminals. Measurements were made using 30-μm thick vibratome sections through the (a) intact, (b) contralateral to the hypoglossal-hypoglossal anastomosis (HHA), and lesioned hypoglossal nucleus either (c) without or (d) with manual stimulation (MS). (e) Graphical representations of the intensity of fluorescence in the intact and lesioned hypoglos-sal nucleus after immunostaing for synaptophysin and Cy3 as florescent dye. Sections were photographed at ×16 magnification. Shown are mean values ± SD of pixel numbers within the defined range of gray values (51-210). Each experimental group comprised 8 rats. The asterisk indicates a significant reduction in the number of pixels in the HHA-only group when compared with those of the intact rats group and the HHA + MS group according to 1-way analysis of variance with the post hoc Bonferroni test.
acetylcholine receptors in motor endplates were stained with Alexa Fluor 488-conjugated bungarotoxin (1:1000, Molecular Probes, Carlsbad, California). All axonal branches that entered or left the boundaries of individual endplates were counted (Figure 8a,b). Counts were per-formed directly under the microscope (objective ×40) in a blind fashion. Preterminal branches of 1 axon were counted as single events and identified as “monoinner-vated.” “Polyinnervated” endplates had 2 or more axons; denervated endplates had no visible axons associated with the receptor staining. We adopted the term polyinnervated to reflect similarity to morphological abnormalities in skeletal muscle of adult mammals following nerve damage or intoxication. In these instances, axonal branching occurs, which may be collateral (at nodes of Ranvier), ter-minal (from endplate terter-minals), or both,21with the result
that many individual endplates are innervated by more than 1 motoneuron (ie, they are “polyneuronally” inner-vated). Our use of the term polyinnervated, rather than
polyneuronally innervated, endplates indicates that we
focused on the axons themselves rather than the perikaryal origins of supernumerary axons in individual endplates (ie, we did not identify 1 or more parent motoneurons).
Estimation of Cortical Tongue Muscle
Representation Volume
Cortical motor representation of musculature has been visualized previously using c-Fos immunoreactiv-ity that is upregulated after axotomy and is a marker for transsynaptic neuronal activation.22
Figure 7. Schematic drawings demonstrating an overview of the supra- and infrahyoid musculature of the (a) rat and (b) detailed localization of the hyoglossus muscle. Whereas a dif-ferentiation among m. geniohyoideus, m. mylohyoideus, m. sternohyoideus, and m. omohyoideus is sometimes hard to achieve, this thin muscle sheet (arrow) extends unvariably from the hyoid bone and enters the tongue laterally, between the masseter and stylohyoideus muscles. Adapted from Greene EC. Anatomy of the rat. Trans Am Philos Soc. 1935;27:1-370.
Figure 8. Motor endplates in (a) intact and (b) reinnervated rat hyoglossus muscle visualized by staining of the motor end-plates with Alexa Fluor 488 α-bungarotoxin (green fluores-cence) and immunostaining of the intramuscular axons for neuronal class III β-tubulin (Cy3 red fluorescence). The abun-dant intramuscular axonal sprouts (in panel b, but not in panel a) cause polyinnervation of the endplates. This is demonstrated in panel b, where each of the 3 motor endplates (indicated by arrows) is reached by at least 3 axons. In con-trast, the motor endplates in the intact hyoglossus (a) are monoinnervated (ie, reached by 1 single axon) (arrows).
Figure 9. Quantification of cortical tongue muscle representation volume. (a) Ventral aspect of the rat cerebrum with depicted rostral borderline of the brain slice (ie, the rhinal fissure; bregma 5.0 mm). Photograph from Figure 8 of Zeman W, Innes JRM,
Craigies Neuroanatomy of the Rat. New York: Academic Press; 1963:23. Reprinted with permission. (b) Schematic drawing of the rat
brain indicating the dimensions of the slice containing the tongue motor area (adapted from Donoghue JP, Wise SP. Rat motor cor-tex: cytoarchitecture and microstimulation mapping. J Comp Neurol. 1982;212:76-88). The entire brain was steadily positioned in a rat brain matrix (RBMS-300C, World Precision Instruments), (c) which allowed identical cutting for all animals’ slice through the telencephalon. A demarcation of the side contralateral to the hypoglossal-hypoglossal anastomosis (HHA) (left) was made using a canule-perforation in the caudoputamen (d). The cortical representation of the suprahyoid-sublingual region (intact or reinner-vated) is in the anterior-lateral neocortex TM1 of both cerebral hemispheres (e, f), as identified by c-Fos immunoreactive neurons (with intranuclear localization of the DAB-HRP immunoreaction product) 1 hour after transection of the right hypoglossal nerve. The portions containing reactive cortical motoneurons were delineated, their areas calculated, and the volume determined accord-ing to the Cavalieri principle.
Tissue preparation. In both intact and operated animals,
the hypoglossal nerve was transected and animals per-fused with fixative 1 hour later. The clearly visible ventral rhinal fissure (bregma 5.0 mm; Figure 9a,b) was selected as the rostral border of the cortical region containing the tongue-muscle motor area. The brain was securely posi-tioned in a rat brain matrix (RBMS-300C, World Precision Instruments, Berlin, Germany; Figure 9c), allowing iden-tical slices to be cut through the telencephalon with a razor blade in each animal. A demarcation of the side contralateral to the HHA (left) was made using a canule-perforation in the caudoputamen (Figure 9d). The cry-oprotected (sucrose-infiltrated) slice was cut coronally (100 sections; 50 μm thick) and sections mounted on SUPERFROST/Plus slides (Roth, Germany).
Immunohistochemical staining. c-Fos was detected using
rabbit antihuman cFos-Ab-5 (1:5000; PC38, Merck Biosciences, Nottingham, UK), biotinylated antirabbit IgG (DakoCytomation, Hamburg, Germany), streptavidin– horseradish peroxidase (HRP) conjugate (1:100; Amersham, Freiburg, Germany), and 3,3′-diaminobenzidine tetrahy-drochloride (Sigma). All sections used to compare the cor-tical tongue muscles representation volume (CTMRV) between stimulated and nonstimulated rats were incu-bated simultaneously using identical solutions.
Quantification of the reactive cortex volume. Using the
fractionator sampling strategy,23each 10th coronal
sec-tion (a total of at least 10 equidistant secsec-tions through the brain) was used for immunocytochemistry of c-Fos. A Zeiss microscope equipped with a CCD Video Camera System (Optronics Engineering Model DEI-470, Goleta, California, supplied by Visitron Systems, Puchheim, Germany) combined with Image-Pro Plus 5.0 software (Media Cybernetics) was used to quantify the projection areas (μm2) containing c-Fos-positive neurons
in each section at a primary magnification of ×2.5. Cortical tongue muscle representation volume was calcu-lated according to the Cavalieri method.23Measurements
were performed by 3 observers (P. Schweigert, S. K. Angelova, and D. N. Angelov) who had no information about treatment of the rats.
Statistical Analyses
Both parametric (2-sided t test for independent sam-ples and 1-way analysis of variance [ANOVA] with post hoc Bonferroni’s test for multiple comparisons of group mean values) and nonparametric tests (Mann-Whitney test) were used to analyze the data as appropriate. The threshold probability value for acceptance of differences was 5%. Throughout the text and in the figures, data are
represented as mean group ± standard deviation. Statistica 6.0 software (StatSoft, Tulsa, Oklahoma) was used for analysis.
RESULTS
Beneficial Effect of Manual Stimulation on
Restoration of Tongue Position
Daily MS of the suprahyoid-sublingual region 5 minutes a day for 2 months improved tongue position after HHA. Deviation of the tongue tip following MS was significantly lower compared with nonstimulated animals (37.4 ± 9.37° vs 50.1 ± 9.01°; mean ± SD, n = 26, Mann-Whitney test, P= .026).
No Effect of Manual Stimulation on
Collateral Axonal Branching
In intact rats, retrograde labeling of the medial hypoglossal branch, which innervates the tongue pro-truders (m. genioglossus) and the intrinsic suprahyoid-sublingual region (vertical and transverse), revealed 2038 ± 1057 (mean ± SD; n = 8) FB-labeled perikarya localized within the ventral hypoglossal subnucleus. Labeling of the lateral branch revealed 835 ± 478 FG-labeled perikarya located in the dorsal hypoglossal sub-nucleus (Figure 3b). No double-labeled perikarya were observed (ie, the index of axonal branching was 0%).
After HHA, 2 major changes, characteristic of aberrant reinnervation of targets, were detected. First, the hypoglos-sal nucleus lacked somatotopy—that is, perikarya were not organized into a dorsal subnucleus (with axons projecting into the lateral hypoglossal nerve branch) and ventral subnucleus (axons projecting into the medial branch of the hypoglossal nerve).24-26 Loss of somatotopy (Figure
4a,b) was due to transection of the hypoglossal nerve proximally to its bifurcation into its medial and lateral branches and subsequent inaccurate navigation of regrowing neuritis into inappropriate branches.
Second, the entire hypoglossal nucleus contained double-labeled (FB + FG) motoneuronal perikarya (arrows in Figure 4a,b), which arose due to collateral axonal branching at the lesion site. Following MS, there were 1234 ± 592 FBonly, 313 ± 460 FGonly, and 1302 ± 715 FB + FG double-labeled perikarya. In nonstimulated rats, there were 600 ± 364 FBonly, 632 ± 722 FGonly, and 1616 ± 691 FB + FG double-labeled perikarya (mean ± SD; n = 8). The index of collateral axonal branching did not differ between the MS and non-MS groups (46% vs 56%; Mann-Whitney test, P> .05).
Increased Size of the Hypoglossal
Motoneurons After Manual Stimulation
Two months after transection and suture of the right hypoglossal nerve and 10 days after surgery for retro-grade labeling, back-labeled motoneuron perikarya in rats receiving MS were significantly larger than in non-stimulated animals (696 ± 53 μm2 vs 594 ± 52 μm2,
group mean values from individual animal mean val-ues, P< .05, t test). This conclusion was further verified by analysis of the frequency distributions in the 2 pop-ulation samples (Figure 5).
Manual Stimulation Restores Normal
Levels of Total Synaptic Input in the
Hypoglossal Nucleus Following
Hypoglossal Nerve Injury
To assess changes in total synaptic input to the hypoglossal nucleus in the 3 groups (Table 1), we quanti-fied synaptophysin expression in 10 equidistant sections at ×16 magnification. Immunocytochemical staining with 1:4000 antisynaptophysin revealed numerous small immunoreactive puncta within the neuropil of the hypoglossal nucleus (Figure 6a-d).
The intensity of synaptophysin immunofluorescence differed significantly between the groups (n = 8 rats in each). Although there were no differences in the pixel distribution curves (data not shown), statistical analysis of total pixel numbers (gray values 51–210) revealed that MS restored synaptophysin levels to those in normal intact animals (Figure 6d). Specifically, compared with intact animals (group 1), synaptophysin levels were restored in animals receiving MS (group 3) but remained lower in those without MS (group 2; P= .022); likewise, following HHA, synaptophysin levels were sig-nificantly higher (P = .011) in animals receiving MS (group 3: HHA + MS) compared with those that did not (group 2: HHA only).
Manual Stimulation Improved the Quality
of Motor Endplate Reinnervation
The number of poly- and monoinnervated motor endplates in the hyoglossus muscle of rats receiving MS was respectively 327 ± 126 and 1464 ± 289; in nonstim-ulated rats, numbers were 375 ± 151 and 803 ± 329. The proportion of polyinnervated endplates was signifi-cantly less in rats receiving MS versus nonstimulated animals (18% vs 32%; ANOVA and post hoc Bonferroni test, P< .0001).
No Effect of Manual Stimulation
on the Dynamic Reorganization
of the Motor Cortex
One hour after transection of the right intact
hypoglossal nerve (group 4), c-Fos immunopositive
nerve cells were seen in the tongue muscle projection area in both the left and right anterior-lateral neocortex, with our findings being in agreement with previous studies.27-29Right and left cortical tongue muscle
repre-sentation volumes did not differ, being 0.063 ± 0.02 mm3and 0.052 ± 0.02 mm3 (P< .05), respectively.
At 2 months after HHA, we retransected the right hypoglossal nerve trunk in groups 5 and 6, inducing acute, within an hour, transsynaptic upregulation of c-Fos expression in the motor cerebral cortex.22,30,31As for intact
animals, left and right cortical tongue muscle representa-tion volumes did not differ in either MS (left: 0.07 ± 0.03 mm3; right: 0.08 ± 0.04 mm3; P> .05) or nonstimulated
rats (left: 0.12 ± 0.06 mm3; right: 0.12 ± 0.05 mm3; P>
.05). Furthermore, compared with intact rats, values did not differ following MS or in the nonstimulated animals.
DISCUSSION
We show that brief, daily MS of the suprahyoid-sub-lingual region significantly improves postoperative recovery of tongue position. Before discussing the potential of these findings for enhanced rehabilitation strategies in detail, we would like to address 3 important critical remarks.
The first one is whether we can provide sound evi-dence that the observed differences in the pattern of rein-nervation and tongue-tip deviation represent an endpoint and not just a delayed recovery in the nonstimulated ani-mals. Of course, we cannot exclude that a slight improve-ment might take place in these animals. However, based on several long-term clinical data and our own observa-tions, we anticipate exactly the opposite. The persisting polyinnervation of muscles (in contrast to the events dur-ing embryonic development, in which there occurs no retrieval of excessive axonal terminals from the neuro-muscular junctions in adults) causes a wasting palsy (paresis and atrophy of the muscles) and progressive func-tional deterioration.32Anyway, to clear this issue definitely,
new experiments with longer postoperative survival peri-ods have to be performed.
A second possible criticism could be that the stimu-lated rats have been subjected to presumably pleasant rather than neutral attention, including gentle stroking followed by a honey “reward.” In this way, it might appear that the stimulated rats spent their 2 postopera-tive months under somewhat “enriched” environmental
conditions compared with the controls. Of course, stim-ulated rats have been dwelling under conditions of “enriched environment” and special “handling” by humans. Our experience with manual stimulation of the vibrissal muscles after facial nerve reconstruction clearly shows that these 2 procedures have no influence on the pattern of reinnervation and recovery of function.10
Finally, it is well known that frequent “handling” of rodents by human beings may trigger multiple neu-roendocrine processes, which could seriously jeopardize our present conclusions. The handling procedure itself is considered as providing a stressful or enhanced envi-ronment, both of which have been shown to affect neu-rogenesis in adult rodents. Hence, the performance of a control with handling but no massage is obligatory. We abandoned this control in the present study only because our earlier work with MS on the rat vibrissal muscles showed definitely no effect of handling.10
Clinical Relevance of Peripheral
Hypoglossal Nerve Injury
Rapid and accurately adjusted tongue movements are paramount for a wide range of functions, including breathing, swallowing, licking/mastication, gaping, gag-ging, coughing, sneezing, vocalization, and vomiting.15,33,34
Although hypoglossal nerve injury has anecdotally been considered rare, lesions may result from tumors,35
trauma,36tonsillectomy,37anterior cervical spinal surgery,38
orotracheal intubation,39 carotid end-arterectomy,40,41
and use as donor tissue for facial reanimation surgery.16
Unilateral hypoglossal damage is considered clinically to be well tolerated due to preservation of taste and tac-tile sensitivity. Furthermore, despite progressive tongue atrophy, only about 10% of patients report difficulties in chewing, swallowing, and speaking at a 6-month follow-up; however, between 6 and 12 months after damage, dysarthria and dysphagia may dramatically worsen, dysfunction that is due to an ongoing aberrant reinnervation.42 Similar to the face, the tongue
com-prises many muscles that, although innervated solely by the hypoglossal nerve, often have antagonistic actions.16
Indeed, in laboratory animals, surgical hypoglossal nerve repair did not result in functional recovery due to aberrant axon regrowth and a failure to reach appropri-ate target muscles.43
Therapeutic Strategies After Peripheral
Nerve Injury
Clinically, there are few options for treating dener-vated muscles, and restoration of useful function after
peripheral nerve injury is a major challenge for recon-structive surgery and rehabilitation medicine.44
Electrical stimulation (ES) has been used therapeuti-cally, although there is a great deal of controversy with either some small benefit being reported or no effect.45
One possible benefit of ES, however, might be that it inhibits intramuscular sprouting and diminishes motor endplate polyinnervation.8,46
Nevertheless, for muscle fibers that are totally dener-vated, regular ES has 2 detrimental effects. First, it sup-presses the production of chemical mediators required for reconnection of an axon branch with its motor end-plate; second, it reduces spontaneous electrical activity within the orphaned muscle fibers (fibrillation), which can act as a signal to induce sprouting in any healthy motor nerves that might remain.47,48Similarly, for muscle
fibers that retain a partial nerve supply, ES may stimulate voluntary muscle overuse and contribute to the suppres-sion of chemical mediators required for reinnervation.49
Other major factors to take into consideration are the changes that occur within the muscle tissue itself. Following denervation and before reinnervation, mus-cles are rapidly and severely affected by loss of bulk and circulation as well as connective tissue shrinkage and fibrosis.50Long-term changes following several months
of complete denervation also include muscle mem-branes becoming relatively nonresponsive to electrical stimulation. In cases where patients are expected to experience nerve regrowth after complete denervation, fibrosis of the muscle connective tissue must be mini-mized to preserve movable muscle structures; in this way, reacquisition of the contractile proteins that make muscles work can occur after muscle reinnervation.51
By contrast, soft tissue massage, tongue-muscle training, and tactile stimulation are thought to promote muscle blood flow and maintain optimum condition during hypoglossal axon regrowth and have shown pos-itive benefits clinically.9,52,53We therefore decided to try
a novel approach and use manual stimulation after hypoglossal nerve injury, gently stroking the extrinsic suprahyoid-sublingual region by hand for 5 minutes daily for 2 months after HHA surgery.
Possible Mechanisms Underlying Improved
Function Following Manual Stimulation
Mechanisms limiting functional recovery after peripheral nerve injury are poorly understood. Our model provides unique opportunities to investigate the influence of MS on both structure and function.
Cortical representation. Cortical reorganization has been
within cortical and subcortical networks is thought to be involved in clinical examples of muscle reanima-tion.55 One reason for finding no changes in cortical
representation is possibly that we examined animals at 2 months, that is, once target reinnervation had been completed. Examination of animals at earlier stages would determine whether cortical representation was altered during reinnervation. Another possibility is that, in our HHA model, sensory innervation of the tongue musculature remains intact with no alteration in somatosensory cortical representation. The lack of quantitative changes in the motor cortical representa-tion regardless of MS supports the norepresenta-tion that cortical plasticity may be of primary importance for restoration of primarily sensory, but not motor, function.22
Collateral branching of axons at the lesion site. Collateral
axonal branching and regrowth to incorrect muscles was not affected by MS. The finding is surprising given that axonal branching and subsequent misdirection of collateral branches have been considered major factors limiting recovery of mimic muscles.3,18Short-term
elec-trical stimulation of the proximal nerve stump immedi-ately after injury has been shown to improve the speed and accuracy of reinnervation in the rat femoral nerve paradigm,56but the functional impact of this treatment
is not known.
However, MS led to improved recovery of tongue position, whereas function was poor to nonexistent in the absence of MS. Thus, as previously shown,10robust
collateral axonal branching appears to be an invariant response to axotomy, regardless of manual stimulation. Finally, reduction of collateral axonal branching using antibodies to a variety of nerve growth factors still does not lead to improved function.7 Taken together, these
findings suggest that even if the normally robust exten-sive sprouting can be limited pharmacologically, func-tional benefit is not conferred, thus suggesting that other possible mechanisms are involved. One possibility is use-dependent plasticity in the central nervous system: the recruitment of aberrantly innervating motoneurons may be reduced or modified by spinal reflex mechanisms, and supraspinal control circuitries thereby override and/or minimize functional distur-bances due to collateral branching. Some studies in humans support this possibility. Motoneurons can be “reeducated” to subserve new functions after muscle tendon transfer.57
Motoneuron size. The finding of larger retrogradely
labeled motoneuron perikarya in stimulated than in nonstimulated rats indicates an effect of the MS on regenerated motoneurons. It is possible that MS allevi-ates axotomy-induced motoneuron atrophy during the recovery period. Alternatively, prevention of atrophy
may result from MS inducing a vigorous regenerative response to the second axotomy performed for retro-grade labeling.19We suggest that the larger cell bodies of
stimulated motoneurons indicate a better functional state. Indeed, there is a correlation between degree of recovery and soma size of retrogradely labeled motoneu-rons reinnervating the quadriceps muscle 3 months after transection and repair of the femoral nerve.20
Although we do not know the mechanism whereby motoneuron size is enhanced, we speculate that MS positively influences regenerating motoneurons via enhanced sensory input (see below).
Synaptic input to the hypoglossal motoneurons. Synaptophysin
is an integral membrane glycoprotein of small presynaptic vesicles and neuroendocrine granules.58 Staining with
synaptophysin antibody detects only presynaptic terminals filled with small vesicles but not depleted boutons. We adopted this immunohistochemical approach in our model to examine the well-established rapid detachment of synaptic terminals from motoneurons following nerve injury, a synaptic stripping that is nevertheless reversible. We are thus confident that our results show a decline in sensory input to the facial motor nucleus after facial nerve injury and that manual stimulation restores such input to normal levels.
Motor endplate reinnervation. Another possible
explana-tion for the better restoraexplana-tion of funcexplana-tion after MS is the reduction of the number of polyinnervated muscle fibers. By contrast, in operated rats without MS, there was an abnormally dense meshwork of intramuscular axonal branches and polyinnervated endplates. Polyinnervation of motor endplates has previously been considered as a factor limiting restoration of function.4
Furthermore, we have previously shown that, compared with normal animals, blind rats recovered vibrissal motor function following facial nerve injury, presum-ably due to forced whisker use; furthermore, whisking in blind rats was associated with a reduction in the pro-portion of polyinnervated motor endplates.7The effect
of MS that we show here for hypoglossal and following facial nerve10injury can be explained by previous
stud-ies showing that imposing muscle activity artificially during synaptic formation and consolidation leads to reduction of intramuscular sprouting.46
Sensory input. In the nerve lesion paradigm used here,
motor axons were lesioned, but circuitry conveying sen-sory information from the tongue to the hypoglossal motoneurons via the trigeminal (V), the glossopharyn-geal (IX), and the superior larynglossopharyn-geal (ie, the vagus [X]) nerve was intact.15One possibility is that MS resulted in
recovery of tongue position via enhanced sensory input. After complete spinal cord transection, in which both
motor and sensory function is lost, motoneurons distal to the injury undergo atrophy, and their dendritic trees shrink and become partially deafferented.59Cell atrophy
and deafferentation of motoneurons is also observed after peripheral axotomy, but in contrast to spinal cord injury, these changes could be reversed after target rein-nervation.19,60We speculate that increased sensory input
in stimulated animals may aid the regenerative response of the injured motoneurons via, for example, stimulat-ing plasticity in the brainstem.
In conclusion, whereas the precise mechanisms linking MS, polyinnervation, and restoration of the muscle func-tion are still unknown, hypoglossal nerve injury provides a further clinically relevant model to address this issue. Here we show that mechanical stimulation of the denervated extrinsic suprahyoid-sublingual region and intrinsic suprahyoid and sublingual muscles (m. longitudinalis sup., m. longitudinalis inf. m. transverses, and m. verticalis) can at least partially “override” the negative effects of extensive but inappropriate axonal regrowth in target muscles. The end result is a reduction in the degree of polyinnervation, which in turn significantly improves motor function of the tongue. Our findings have implications for rehabilitation strategies following peripheral nerve injury involving only motor axons because they suggest that therapies should be directed toward muscle reinnervation (ie, within periph-eral target tissue) rather than misdirected axonal regrowth at the site of nerve injury.
ACKNOWLEDGMENTS
The present study was supported by the Partnerschaftsprogramm der Universität zu Köln mit der St. Kliment Ohridski-Universität Sofia, Bulgaria; the Jean-Uhrmacher Foundation; Köln Fortune Programm; and the DFG (AN 331/3-1, AN 331/5-1). SD is a senior research fellow (National Health & Medical Research Council, Australia; grant ID 254670). The skillful assis-tance of D. Bösel, D. Felder, I. Rohrmann, and L. Wilken is highly appreciated.
REFERENCES
1. Kerrebijn JDF, Freeman JL. Facial nerve reconstruction: outcome and features. J Otolaryngol. 1998;27:183-186.
2. Baker RS, Stava MW, Nelson KR, et al. Aberrant reinnervation of facial musculature in a subhuman primate: a correlative analysis of eyelid kinematics, muscle synkinesis, and motoneuron local-ization. Neurology. 1994;44:2165-2173.
3. Ito M, Kudo M. Reinnervation by axon collaterals from single facial motoneurons to multiple targets following axotomy in the adult guinea pig. Acta Anat. 1994;151:124-130.
4. Tam SL, Gordon T. Mechanisms controlling axonal sprouting at the neuromuscular junction. J Neurocytol. 2003;32:961-974.
5. Tomov TL, Guntinas-Lichius O, Grosheva M, et al. An example of neural plasticity evoked by putative behavioral demand and early use of vibrissal hairs after facial nerve transection. Exp
Neurol. 2002;178:207-218.
6. Streppel M, Azzolin N, Dohm S, et al. Focal application of neu-tralizing antibodies to soluble neurotrophic factors reduces col-lateral axonal branching after peripheral nerve lesion. Eur J
Neurosci. 2002;15:1327-1342.
7. Guntinas-Lichius O, Irintchev A, Streppel M, et al. Factors limit-ing motor recovery after facial nerve transection in the rat: com-bined structural and functional analyses. Eur J Neurosci. 2005;21:391-402.
8. Love FM, Son YJ, Thompson WJ. Activity alters muscle reinner-vation and terminal sprouting by reducing the number of Schwann cell pathways that grow to link synaptic sites. J
Neurobiol. 2003;54:566-576.
9. Coulson SE. Physiotherapy rehabilitation following facial nerve paresis. In: Beurskens CHG, van Gelder RS, Heymans PG, Manni JJ, Nicolai JPA, eds. The Facial Palsies: Complementary Approaches. Utrecht, The Netherlands: Lemma Publishers; 2005:263-274. 10. Angelov DN, Ceynowa M, Guntinas-Lichius O, et al. Mechanical
stimulation of paralyzed vibrissal muscles following facial nerve injury in adult rat promotes full recovery of whisking. Neurobiol
Dis. 2007;26:229-242.
11. Svensson P, Romaniello A, Arendt-Nielsen L, et al. Plasticity in corticomotor control of the human tongue musculature induced by tongue task-training. Exp Brain Res. 2003;152:42-51. 12. Yu WHA, Yu MC. Acceleration of the regeneration of the crushed
hypoglossal nerve by testosterone. Exp Neurol. 1983;80:349-360. 13. Sanes JN, Suner S, Lando JF, et al. Rapid reorganization of adult rat motor cortex somatic representation patterns after motor nerve injury. Proc Natl Acad Sci USA. 1988;85:2003-2007. 14. Svensson P, Romaniello A, Wang K, et al. One hour of tongue-task
training is associated with plasticity in corticomotor control of the human tongue musculature. Exp Brain Res. 2006;173:165-173. 15. Lowe AA. The neural regulation of tongue movements. Progr
Neurobiol. 1981;15:295-344.
16. Wilson JR, Sumner AJ, Eichelman J. Aberrant reinnervation fol-lowing hypoglossal nerve damage. Muscle Nerve. 1994;17:931-935. 17. Angelov DN, Neiss WF, Gunkel A, et al. Axotomy induces intranuclear immunolocalization of neuron-specific enolase in facial and hypoglossal motoneurons of the rat. J Neurocytol. 1994;23:218-233.
18. Dohm S, Streppel M, Guntinas-Lichius O, et al. Local application of extracellular matrix proteins fail top reduce the number of axonal branches after varying reconstructive surgery on rat facial nerve. Restor Neurol Neurosci. 2000;16:117-126.
19. McPhail LT, Fernandes KJ, Chan CC, et al. Axonal reinjury reveals the survival and re-expression of regeneration-associated genes in chronically axotomized adult mouse motoneurons. Exp
Neurol. 2004;188:331-340.
20. Simova O, Irintchev A, Mehanna A, et al. Carbohydrate mimics promote functional recovery after peripheral nerve repair. Ann
Neurol. 2006;60:430-437.
21. Son YJ, Trachtenberg JT, Thompson WJ. Schwann cells induce and guide sprouting and reinnervation of neuromuscular junc-tions. Trends Neurosci. 1996;19:280-285.
22. Bisler S, Schleicher A, Gass P, et al. Expression of c-Fos, ICER, Krox-24 and JunB in the whisker-to-barrel pathway of rats: time course of induction upon whisker stimulation by tactile explo-ration of an enriched environment. J Chem Neuroanat. 2002;23:187-198.
23. Gundersen HJG, Bendtsen TF, Korbo L, et al. Some new, simple and efficient stereological methods and their use in pathological research and diagnosis. Acta Pathol Microbiol Immunol Scand. 1988;96:379-394.
24. Krammer EB, Rath T, Lischka MF. Somatotopic organization of the hypoglossal nucleus: a HRP study in the rat. Brain Res. 1979;170:533-537.
25. Uemura-Sumi M, Itoh M, Mizuno N. The distribution of hypoglossal motoneurons in the dog, rabbit and rat. Anat
Embryol. 1988;177:389-394.
26. Aldes LD. Subcompartmental organization of the ventral (pro-trusor) compartment in the hypoglossal nucleus of the rat. J
Comp Neurol. 1995;353:89-108.
27. Rodel RM, Tergau F, Markus H, et al. Bilateral changes in the cor-tical motor representation of the tongue after unilateral periph-eral facial paralysis: evidence from transcranial magnetic stimulation. Ann Otol Rhinol Laryngol. 2004;113:951-955. 28. Donoghue JP, Wise SP. Rat motor cortex: cytoarchitecture and
microstimulation mapping. J Comp Neurol. 1982;212:76-88. 29. Neafsey EJ, Bold EL, Haas G, et al. The organization of the rat
motor cortex: a microstimulation mapping study. Brain Res Rev. 1986;11:77-96.
30. Narita M, Ozaki S, Narita M, et al. Change in the expression of c-fos in the brain following sciatic nerve ligation. Neurosci Lett. 2003;352:231-233.
31. Peeva GP, Angelova SK, Guntinas-Lichius O, et al. Improved out-come of facial nerve repair in rats is associated with enhanced regenerative response of motoneurons and augmented neocorti-cal plasticity. Eur J Neurosci. 2006;24:2152-2162.
32. Greulich M. Anchoring the nasolabial fold. In: Beurskens CH, van Gelder RS, Heymans PG, Manni JJ, Nicolai JPA, eds. The
Facial Palsies: Complementary Approaches. Utrecht, The Netherlands: Lemma Publishers; 2005:235-243.
33. Sawczuk A, Moser KM. Neural control of tongue movement with respect to respiration and swallowing. Crit Rev Oral Biol Med. 2001;12:18-37.
34. Miller AJ. Oral and pharyngeal reflexes in the mammalian ner-vous system: their diverse range in complexity and the pivotal role of the tongue. Crit Rev Oral Biol Med. 2002;13:409-425. 35. Keane JR. Twelfth-nerve palsy: analysis of 100 cases. Arch Neurol.
1996;53:561-566.
36. Brennan RJ, Shirley JP, Compton JS. Bilateral hypoglossal nerve palsy following head injury. J Emerg Med. 1993;11:167-168. 37. Sharp CM, Borg HK, Kishore A, et al. Hypoglossal nerve
paraly-sis following tonsillectomy. J Laryngol Otol. 2002;116:389-391. 38. Sengupta DK, Grevitt MP, Mehdian SM. Hypoglossal nerve
injury as a complication of anterior surgery to the upper cervical spine. Eur Spine. 1999;8:78-80.
39. Rubio-Nazabal E, Marey-Lopez J, Lopez-Facal S, et al. Isolated bilateral paralysis of the hypoglossal nerve after transoral intuba-tion for general anesthesia. Anesthesiology. 2002;96:245-247. 40. Lindsay FW, Mullin D, Keefe MA. Subacute hypoglossal nerve
paresis with internal carotid artery dissection. Laryngoscope. 2003;113:1530-1533.
41. Cunningham EJ, Bond R, Mayberg MR, et al. Risk of persistent cranial nerve injury after carotid endarterectomy. J Neurosurg. 2004;101:445-448.
42. Conley JJ, Baker D. Hypoglossal-facial nerve anastomosis for rein-nervation of the paralysed face. Plast Reconstr Surg. 1979;63:63-72. 43. Hosemann W, Wigand ME, Herrlinger P, et al. Surgical reneuro-tization of the tongue after unilateral paralysis of the hypoglossal nerve. Eur Arch Otorhinolaryngol. 1990;248:95-98.
44. Lundborg G, Richard P. Bunge memorial lecture. Nerve injury and repair: a challenge to the plastic brain. J Peripher Nerv Syst. 2003;8:209-226.
45. Diels HJ. Current concepts in non-surgical facial nerve rehabili-tation. In: Beurskens CHG, van Gelder RS, Heymans PG, Manni JJ, Nicolai JPA, eds. The Facial Palsies: Complementary
Approaches. Utrecht, The Netherlands: Lemma Publishers;
2005:275-283.
46. Brown MC, Holland RL, Ironton R. Nodal and terminal sprout-ing from motor nerves. J Physiol. 1980;306:493-510.
47. Cohan CS, Kater SB. Suppression of neurite elongation and growth cone motility by electrical activity. Science. 1986;232: 1638-1640. 48. Brown MC, Holland RL. A central role for denervated tissues in
causing nerve sprouting. Nature. 1979;282:724-726.
49. Diels HJ. New concepts in nonsurgical facial nerve rehabilitation. In: Myers E, Bluestone C, eds. Advances in Otolaryngology: Head
and Neck Surgery. Chicago: Mosby-Year Book; 1995:289-313.
50. Bardosi A, Goebel HH, Stennert E. The ultrastructure of normal and denervated human facial muscles. Plast Reconstruct Surg. 1987;79:171-176.
51. Nix WA. Effects of intermittent high frequency electrical stimu-lation on denervated EDL muscle of rabbit. Muscle Nerve. 1990; 13:580-585.
52. Hovind H, Nielsen S. Effect of massage on blood flow in skeletal muscle. Scand J Rehab Med. 1974;6:74-77.
53. Lamm NCEdD, De Felice A, Cargan A. Effect of tactile stimula-tion on lingual motor funcstimula-tion in pediatric lingual dysphagia.
Dysphagia. 2005;20:311-324.
54. Horvath S, Prandovszky E, Pankotai E, et al. Use of recombinant pseudorabies virus to analyze motor cortical reorganization after unilateral facial denervation. Cerebr Cortex. 2005;15:378-384. 55. Sanes JN, Donoghue JP. Plasticity and primary motor cortex.
Ann Rev Neurosci. 2000;23:393-415.
56. Al-Majed AA, Newmann CM, Brushart TM, et al. Brief electrical stimulation promotes the speed and accuracy of motor axonal regeneration. J Neurosci. 2000;20:2602-2608.
57. Wiedemann E, Eggert C, Illert M, et al. Functional electromyog-raphy analysis of radial replacement operation. Orthopade. 1997;26:673-683.
58. Wiedenmann B, Franke WW. Identification and localization of synaptophysin, an integral membrane glycoprotein of Mr 38.000 characteristic of presynaptic vesicles. Cell. 1985;41: 1017-1028.
59. Kitzman P. Alteration in axial motoneuron morphology in the spinal cord of the injured spastic rat. Exp Neurol. 2005;192:100-108. 60. Sumner BE, Watson WE. Retraction and expansion of the den-dritic tree of motor neurons of adult rats induced in vivo. Nature. 1971;233:273-275.
