| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Journal of Physiology (2001), 532.3, pp. 835-849
© Copyright 2001 The Physiological Society
 |
ABSTRACT |
|---|
 TopAbstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Peripheral nerve fibres decline in diameter and conduction velocity slows when the fibres are disconnected from their targets. Diameter and conduction velocity recover fully when the cut axons regenerate and reinnervate denervated sensory and muscle targets (Kuno et al. 1974a,b; Gordon & Stein, 1982a,b; Gillespie & Stein, 1983; Gordon, 1983; Foehring et al. 1986a,b; Titmus & Faber, 1990). These findings show that connection with peripheral end-organs is essential to maintain the normal size of peripheral nerve fibres. Target control of fibre size is exerted at the level of neuronal gene expression since: (1) axonal size varies with the number of neurofilaments in the axonal cytoskeleton (Hoffman et al. 1987) and (2) decline in axonal size parallels the downregulation of neurofilament expression in axotomized neurons (Hoffman et al. 1983, 1987). However, the nature of the target control of axonal size via neurofilament gene expression is still not understood.
There is normally a wide range in fibre size in peripheral nerves, the diameter of the fibres varying with the sensory modality and motoneuronal type. Muscle afferents and efferent fibres comprise the largest myelinated fibres with the fastest conduction velocities; the 
-motoneurons to the intrafusal muscle fibres in muscle spindles and 
temperature and pain sensory fibres are smaller (Boyd & Davey, 1966). Some pain fibres and the autonomic efferent fibres are even smaller and are non-myelinated. Within the myelinated nerve fibre populations, there is a range in nerve fibre diameter and conduction velocity. For example, the diameters and conduction velocities of motor nerve fibres that innervate any one skeletal muscle vary by a factor of 4 and 2, respectively (Boyd & Davey, 1966). Neuronal properties, including nerve fibre size and conduction velocity, differentiate during neonatal development (Vejada et al. 1985). A popular model of neuronal differentiation includes neuronal retrograde transport of specific neurotrophic factors from the innervated targets (Korschling, 1993). Uptake of and responsiveness to the factors require the expression of specific receptors and differential sensitivity of the receptors for the neurotrophic factors. Within a given neuronal population, such as the 
-motoneurons, it is possible that the availability of different amounts of neurotrophic factors to different neurons determines the range in nerve fibre diameter and conduction velocity. In this regard, the size of the target could be a factor in the retrograde regulation of fibre size in neuronal populations.
Indirect evidence to support the hypothesis that the size of the target regulates axonal size in motoneurons in the adult is the consistent finding that muscle target and motor nerve fibre size co-vary. Electrophysiological parameters of nerve fibre size, conduction velocity and action potential amplitude, co-vary with motor unit force and innervation ratio in normal and long-term reinnervated motor units (Gordon & Stein, 1982a,b; Rafuse & Gordon, 1996). In experiments where target size was increased (by partial denervation or inducing hypertrophy by functional overload) to test the hypothesis, increased diameters of myelinated nerve fibres were reported (Edds, 1949, 1950a,b). The effects were small however, and there was no documentation of the target size following partial denervation, namely the number of muscle fibres per motor nerve fibre. In light of more recent findings of a much larger increase in diameter of sympathetic axons after partial denervation (Voyvodic, 1989), we felt that it was important to re-examine the issue of whether an increased target size by partial denervation is associated with increased size and conduction velocity in myelinated nerve fibres.
To test the hypothesis that target size regulates size of myelinated nerve fibres, we partially denervated sensory and muscle targets to promote sprouting and to increase the number of targets per nerve fibre. We measured nerve fibre size electrophysiologically and morphologically under conditions in which we quantified the extent of partial denervation and consequently the increase in the target size.
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Surgery and preparation
A total of 14 adult cats (3-3.5 kg) and three kittens (3-4 weeks) were used in this study with approval from the local ethical committee which operates under the Canadian guidelines for animal experimentation. Five adult cats served as controls. In the remaining nine adult cats and three kittens, partial denervation of triceps surae muscles was performed, under sodium pentobarbital anaesthesia (Somnotol: 30 mg kg-1 I.P.), using sterile surgical procedures. Depth of anaesthesia was monitored by absence of withdrawal and pinna reflexes in response to pinch of the forelimb and ear, respectively. Deep anaesthesia was maintained where necessary by halothane inhalation. The L7 and S1 spinal roots were exposed by first clearing the muscle from the lateral process of either the seventh lumbar or first sacral vertebral bone and cutting the process with rongeurs. Either L7 or S1 spinal root was lifted gently with a fine glass hook and cut distal to the dorsal root ganglia that remained intact. No attempt was made to prevent regeneration of the cut roots. In some operations, bipolar stimulating electrodes were used to stimulate the exposed roots to qualitatively assess their contribution to force production in the triceps surae muscles before transecting either the L7 or S1 root. Cats and kittens were injected postoperatively with buprenorphine (0.01-0.02 mg kg-1 I.P.) and placed on a circulating water heat pad at 37 °C to recover over a 24 h period, under close observation. The animals were housed in large cages during recovery from the operation and during the 3-12 months (mean ± S.E.M. of 6.4 ± 0.9 months) before the final acute experiments. Compensatory responses of the intact root were monitored qualitatively by palpation of the triceps surae muscles during quiet and bipedal standing on the hindlimbs. Progressive increases in muscle force, particularly for muscles which suffered extensive denervation, were readily detectable with this method.
Final acute experiments were carried out under deep surgical anaesthesia (Somnotol 35-40 mg kg-1 I.P and maintained by venous injection). Anaesthesia was maintained at a level at which there was complete absence of withdrawal and pinna reflexes. Cats were prepared for bilateral recording from dorsal and ventral roots that were cut centrally for ventral root splitting (to record from single motor units) and for force recording from the medial gastrocnemius (MG) muscle. A laminectomy from L5 to S2 was performed to allow exposure of the L6, L7, S1 and S2 dorsal and ventral roots. The roots were cut as they entered the spinal cord. MG muscles were exposed and prepared for force recordings by isolating and tying their distal tendon with a silk thread (no. 5). A small piece of calcaneous bone was left attached to the tendons to prevent the suture from slipping. All other hindlimb muscles were denervated by sectioning all the nerves other than the MG branch of the sciatic nerve. MG nerves were freed over a distance of at least 2-3 cm from their entry into the muscles in order to place the nerves over an electrode array of 2-3 cm without stretching the nerve (Fig. 1A).
Charge measurement on dorsal and ventral roots
As described in detail elsewhere (Hoffer et al. 1979), we measured the charge contributed on each root bilaterally by the MG nerve as a means of estimating: (1) the proportion of the motor and sensory nerve fibres which were cut by the unilateral surgical section of one of two spinal roots and (2) the size of the remaining nerve fibres in the uncut roots which supply enlarged targets by putative sprouting. Briefly, each of the exposed dorsal and ventral roots on left and right sides of the spinal cord was sequentially freed and cut centrally for placement on a six electrode recording array with a 2 mm interelectrode distance to measure: (1) the impedance (in kilo-ohms) of each root using a 10 kHz sinusoidal signal and (2) the mean of compound monophasic action potential (CAP; 10-20 mV) responses to supramaximal stimulation of the MG nerves (Fig. 1A). Current recorded on each root was derived by division of the CAP (in millivolts) by the impedance (in kilo-ohms) and charge was obtained by integration of the current traces. Since charge can be summed (in contrast to voltage), and the distribution of axons to each root are bilaterally symmetrical (Hoffer et al. 1979), the proportion of axons cut on one root in the partially denervated side was obtained by the ratio of the charge on the corresponding intact contralateral root to the sum of charges on both the intact contralateral roots. The size of the CAP (in millivolts), measured on each root reflects the number and size of the nerve fibres in the root once the impedance of the root is taken into account by normalizing the CAP by impedence to obtain current (nanoamperes). The CAP is underestimated by up to 10 % by dispersion of the unitary action potentials with different conduction velocities (Hoffer et al. 1979).
Under some conditions, cut spinal roots regenerated. In those cases, charge recorded on the cut ventral and/or dorsal roots provided an estimate of the number and/or size of regenerated nerve fibres, which together with records of evoked muscle forces, provided an estimate of axonal regeneration from the cut spinal root.
Isometric force recording
CAPs and tetanic forces were recorded in response to stimulation of each ventral root as shown in Fig. 1A. The tetanic force, evoked by simultaneous stimulation of the two ventral roots (i.e. L7 and S1) containing motor nerve fibres to the muscle in question, is approximately equal to the additive force elicited upon stimulation of each root separately (Fig. 1C). Similarly, the size of the CAPs, recorded from each root separately upon stimulation of the MG muscle nerve, summate and equal the size of the CAP recorded from both roots simultaneously (Fig. 1B). The relative contribution of nerves in each root can therefore be obtained by the ratio of the force elicited by stimulation of one root to the sum of the forces elicited by all motor nerves in both the contributing ventral roots or the muscle nerve. This measure of the contribution of the ventral root to the muscle nerve was also used to calculate the extent of muscle partial denervation carried out previously by cutting the corresponding root on the contralateral side. The agreement with the value of partial denervation from charge recordings was within 3 %.
As illustrated in Fig. 1, evoked muscle forces in the partially denervated MG muscles were recorded in response to the supramaximal stimulation of the intact ventral roots to provide an estimate of the extent of putative sprouting, or growth of nerve terminals from uncut axons to reinnervate denervated muscle fibres. If, for example, 80 % of the innervation was removed by cutting the S1 spinal root and evoked forces in response to stimulation of L7 ventral root equalled the maximum force in the contralateral normal muscle, putative sprouting from intact motor nerve fibres in the partially denervated muscle had enlarged motor units 5-fold. The extent of putative sprouting was verified by recording evoked forces from isolated motor units in the partially denervated muscles. Stimulation of the cut ventral root was also carried out to determine whether regeneration had occurred. Regeneration refers to axonal growth from the proximal ends of cut nerves. Under conditions in which regeneration had occurred, the evoked forces provided an estimate of the extent of axonal regeneration and muscle reinnervation.
Recording of unitary action potentials and motor unit forces
As described in detail previously (Rafuse et al. 1992; Rafuse & Gordon, 1996), ventral roots were dissected to isolate motor nerve fibres to the partially denervated MG muscles for electrophysiological measurement of nerve fibre and motor unit size. Briefly, ventral roots were dissected until 1 Hz stimulation elicited all or none: (a) unitary action potentials on an array of five electrodes placed on the MG nerve (a triphasic configuration was used with the outermost two electrodes earthed to reduce contamination of the signal from the EMG), (b) EMG responses on bipolar EMG electrodes on a Silastic pad sewn onto the fascia of the muscle and (c) twitch forces recorded using a Grass force transducer (FT 03). Twenty to 30 evoked neural and muscle action potentials and twitch forces were averaged using a PDP 11/21 microcomputer and stored on disk for further analysis. In addition, tetanic force was evoked by stimulation at 100 Hz (20 pulses) and averaged at a rate of 0.5 Hz. Recordings were made from at least ten motor units from each muscle.
Conduction velocity was calculated by dividing the latency of the unitary action potentials by the conduction distance. Conduction velocity and peak-to-peak amplitude of the unitary action potentials were both used as physiological estimates of the size of individual motor nerve fibres which innervated single motor units (Gordon & Stein, 1982a,b). Unitary muscle twitch and tetanic forces were used as indirect estimates of the size of enlarged motor units which reflects the extent of the enlargement of the muscle target per motor nerve fibre. As motor unit force is the product of the number, size and specific force of the muscle fibres supplied by one motor nerve fibre, motor unit force is a reasonable measure of the number of muscle fibres per motor nerve fibre or innervation ratio, provided the size and specific force of the muscle fibres do not change appreciably (see Rafuse et al. 1992; Rafuse & Gordon, 1996).
Immediately following the conclusion of each experiment and removal of nerve and muscle tissue, the cats were killed by an overdose injection of Somnotol (200 mg kg-1).
Nerve histology
At the termination of the electrophysiological recordings, MG nerve segments were removed for histological examination. Segments, 5-10 mm long, were taken: (1) at the site of the electrode placement for stimulation and recording of unitary action potentials and (2) proximal and (3) distal to the electrode site. The nerve segments were fixed in gluteraldehyde (3 % in 0.1 M phosphate buffer), stained with OsO4 (3 % solution in 0.1 M phosphate buffer), dehydrated in ascending alcohols and embedded in Araldite. Cross-sections of 1 mm were mounted, stained (1 % p-phenylenediamine) and microphotographed for measurement (see Gillespie & Stein, 1983; Gordon et al. 1991). Two methods of measuring myelinated nerve fibre size were used. Both methods gave comparable results. Sample areas of nerve cross-sections were either photographed and printed at magnifications of 
1200-2000 for digitization of outer perimeters of myelinated axons, or directly analysed using a PSICOM 327 (microscope-computer linked system) with PSIEXEC software for proprietary image processing (Perceptive Systems Inc., League City, TX, USA). The outer perimeter of the nerve fibre (including axon and myelin sheath) was used to determine the fibre area and in turn, the diameter of an equivalent circle.
Total fibre numbers in the nerve were calculated by counting fibres in 4-8 sample photographs of the nerve cross-section at the same magnification and by multiplying the mean count by the ratio of the area of the whole nerve to the area of the sample counted.
Statistical treatment
Means were tested for significant difference by testing the null hypothesis of variance. Arithmetic means are given with standard errors (S.E.M). Relationships on X-Y plots were fitted with straight lines according to mean squares criteria and the Pearson product-moment correlation coefficients (r). X and Y variables were considered to be correlated if the slope of the regression lines was significantly different from zero with P < 0.05. Student's t test was used to determine differences in the mean size of the myelinated nerve fibres in control and partially denervated limbs. Differences between distributions was determined using the Kolmogorov-Smirnov test (Fisz, 1963).
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Mean values (± S.E.M.) of muscle forces and electrophysiological measurements of charge, unitary action potentials and conduction velocities in adult cats and kittens were not statistically different from each other after partial denervation (Table 1). Hence, the data obtained from the adult cats and kittens are presented together.
Number and size of sensory and motor nerve fibres after partial denervation: charge measurements
Root distributions. Motor and sensory nerve fibres to the MG muscle exit the spinal cord through the L7 and S1 spinal roots since CAPs were recorded only on the L7 and S1 ventral and dorsal roots in response to stimulation of the MG nerve (Fig. 1B). CAPs recorded on each root separately equal the CAP recorded on both roots simultaneously. However, the smaller size of the S1 root (and hence larger impedance) yields a relatively larger voltage difference. CAPs can be compared more directly by dividing voltage by root impedance. Division of voltage (in millivolts) by root impedance (in milliohms) gives current (in nanoamperes), and by integration electrical charge (in picocoulombs), which can then be used to estimate the relative number and size of the nerve fibres in each root (see Methods and Hoffer et al. 1979). As the distribution of nerve fibre size in each root is not different (see Hoffer et al. 1979), the charge for each root, relative to the sum of the charge on both roots, is equal to the relative number of nerve fibres which enter or exit the spinal cord via that root. For example, in Fig. 1B, charge contributions of the L7 and S1 ventral roots of 10.8 and 124.3 pC correspond to the relative proportions of 8 and 92 % of the MG motor nerve fibres that exit the spinal cord via the L7 and S1 ventral roots, respectively.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 1. Experimental methods for recording of monophasic compound action potentials (CAPs) and evoked isometric forces from the MG muscle in response to stimulation of the MG nerve A, schematic diagram of the method used to record CAPs on L7 and S1 dorsal and ventral roots in response to stimulation of MG nerve after chronic unilateral section of either L7 or S1 spinal root, 6.4 ± 0.9 months previously. Nerve branches other than the nerve to the MG muscle were cut. These nerves included the common peroneal (CP) and tibial (Tib) branches of the sciatic (Sc) nerve. MG nerve was stimulated supramaximally to evoke CAPs on the roots as well as to evoke twitch and tetanic contractions. B, the extent of partial denervation was determined from the relative contribution of CAPs recorded from the S1 and L7 ventral root in response to stimulation of MG nerve on the contralateral unoperated side. The CAPs (in millivolts) shown here, were normalized by root impedance (in kilo-ohms) to derive values of current (nanoamperes) and charge (by integration). The charges contributed on each root add to give a signal whose amplitude depends on the size and the total number of motor nerve fibres exiting in the root. C, the isometric tetanic muscle forces recorded in response to stimulation of the L7 and S1 ventral roots provided a complementary estimate of proportion of motor nerve fibres exiting in each root (see text for further details).
| ||
Partial denervation. As the number of nerve fibres that exit L7 and S1 roots are bilaterally symmetrical in any one animal (Hoffer et al. 1979), the relative proportions of nerve fibres that exit a root on one side of the spinal cord gives the relative proportion of fibres in the same root on the other side. Therefore, we used the contralateral charge and force measurements to estimate: (1) the proportion of motor and sensory neurons that were axotomized by transection of one spinal root and (2) the extent of partial denervation of the target muscles. In nine cats, the proportion of motor nerve fibres in the L7 root varied from 8 to 88 % with a mean (± S.E.M.) of 47.0 ± 8.3 % for the MG nerve in the experimental adult cats. The corresponding mean value for the kittens was 64.0 ± 19.6 % (Table 1). Charge and force estimates of relative nerve fibre proportions in the two roots are in good agreement (Table 1).
Uncut nerve fibres with enlarged targets. To test whether there was a retrograde effect of putative axonal sprouting and/or enlargement of target size on nerve fibre size, we cut one of two spinal roots and compared charges (in picocoulombs) on the ipsilateral uncut dorsal and ventral roots (Fig. 2A and B; L7 filled bars) with the charges on the corresponding contralateral intact roots (Fig. 2A and B; L7; open bars). There was no significant difference between the values (Fig. 2A and B). The number and size of nerve fibres exiting the roots are bilaterally the same and transection of one of two spinal roots does not affect the number of nerve fibres in the intact root. Consequently, the same charge measured on nerve fibres that innervate normal and enlarged targets, shows that an enlarged target does not increase the mean size of either motor or sensory nerve fibres.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 2. Charge delivered to intact L7 dorsal (DR) and ventral (VR) roots in response to stimulation of MG nerve is unchanged after cutting S1 spinal root in contrast to the force in the partially denervated MG muscles which is increased Mean (± S.E.M.) charge values delivered by the MG muscle nerves to the L7 and S1 dorsal (A) and ventral roots (B) in nine cats in which S1 spinal root was cut unilaterally 5.5 ± 0.6 months previously (filled bars) and compared with the charge delivered to the roots on the contralateral unoperated side (open bars). C, mean isometric tetanic force elicited by stimulation of the intact L7 and cut S1 roots in partially denervated MG muscles. Note that mean charge contributions for control and partially denervated nerves to the L7 root were not different, in contrast to the tetanic force elicited by L7. Total charge declined in contrast to total force which was not different from normal. Significant differences between means are shown: * P < 0.05.
| ||
A second method of analysis of the charge data support the above results in indicating that enlarged targets do not have a retrograde effect on size of the nerve fibres. The charge measured on the intact ipsilateral L7 roots, after cutting the S1 spinal root, does not equal, or even approach, the arithmetic sum of charges recorded from the two contralateral intact L7 and S1 roots which supply normal targets (Fig. 2A and B; L7 + S1; open bars). The charge deficits of the remaining ventral and dorsal roots equalled the charge 'lost' from the cut S1 roots, as measured from the corresponding root on the contralateral side of the spinal cord (Fig. 2A and B). This data again shows that there was no compensatory increase in charge on the intact L7 root whose axons undergo putative sprouting to reinnervate partially denervated targets. The sprouting is evident by the increased force evoked by stimulation by the intact L7 ventral root. The evoked force equalled the sum of the muscle forces evoked by both ventral roots on the contralateral side (Fig. 2C; L7 + S1; open bars).
Regeneration of cut ventral roots
Stimulation of the MG nerves did not evoke CAPs on the cut ventral or dorsal roots unless regeneration occurred during the 6.4 ± 0.9 months between cutting one spinal root and the acute experiment. The absence of any sensory nerve regeneration was evident by zero charge elicited on the cut S1 dorsal root by stimulation of the MG nerve (Fig. 2A). There were a few animals in which motor regeneration occurred despite complete spinal root transection and no surgical repair. This was detected by the small CAPs recorded on the cut ventral roots in response to stimulation of the peripheral nerves (Fig. 2B). These were detected at periods of 4 months or longer after cutting one spinal root. The small size of the CAPs indicated that the regenerating nerve fibres were small and/or few in number. There were corresponding stimulation-evoked contractions in the partially denervated muscles (Fig. 2C). These were recorded only from those muscles where the spinal root transection produced a partial denervation of > 80 % and recordings were made at least 4 months after cutting the spinal root, consistent with previously observations of Rafuse et al. (1992) that regenerating axons only make functional connections in partially denervated muscles where the partial denervation is > 80 %. Nonetheless, the force contribution of these nerve fibres was small as compared with the force contribution of the nerve fibres in the remaining intact root (Fig. 2C). Thus, regenerated motor nerve fibres from the cut spinal root only contributed a small proportion of the reinnervation of the partially denervated muscles. The axons of the intact neurons were the major source of innervation of the extensively denervated muscles.
Enlarged muscle targets. Increase in the MG muscle target size was estimated from the amplitude of the twitch and tetanic contractions in the partially denervated MG muscles (Fig. 1). Stimulation of the L7 ventral root evoked an average of 2.4 the tetanic force in the partially denervated muscles as compared with the contralateral normal muscles. As shown in Fig. 2C, the increased muscle force compensated fully for the loss of innervation from the cut S1 spinal root: tetanic force elicited by L7 root stimulation equalled the sum of the forces developed in response to L7 and S1 root stimulation in the contralateral hindlimb.
Relative charge and force as a function of partial denervation. The charge on the dorsal and ventral uncut roots declined as a function of the charge loss produced by cutting the other spinal root (Fig. 3A). Regression analysis showed that, when a line was fitted to the data points, the slope (± S.E.M.) of -0.84 ± 0.12 (r = 0.83) was not statistically different from -1 which is the slope predicted for a condition where no change occurred in the size of the nerve fibres in the ipsilateral uncut dorsal and ventral roots. The inverse relationship between charge remaining and percentage partial denervation therefore reflects the reduced number of nerve fibres in the remaining root and constancy of their mean size. Taken together, these data demonstrate that there were no compensatory increases in charge for nerve fibres that supplied enlarged motor units.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 3. Relative charge on dorsal and ventral roots and tetanic force in response to stimulation of the MG nerve after section of either L7 or S1 spinal roots, plotted as a function of the percentage partial denervation A, the charge contribution on the uncut dorsal (
| ||
Muscle forces recorded in the partially denervated muscles, on the other hand, did not decline as numbers of nerve fibres were reduced by spinal root section (Fig. 3B). If the intact nerve fibres did not sprout and innervate denervated muscles fibres to compensate for the reduced nerve fibre number, force would decline in direct proportion to the extent of partial denervation (Fig. 3B; diagonal line). The deviation from the theoretical line reflects the increase in muscle force generated by the reduced numbers of nerve fibres, because each nerve fibre now innervates more muscle fibres. For extensive partial denervation (> 80 %), an upper limit to motor unit size prevented complete compensation by putative sprouting from the intact motor nerves (Fig. 3B ; see also Rafuse et. al. 1992; Rafuse & Gordon, 1996). The tendency for regenerating nerve fibres from the cut spinal root to reinnervate denervated muscle fibres to form small motor units in extensively denervated muscles (> 80 %) accounted for the full recovery of whole muscle force (Fig. 3B; see also Rafuse et al. 1992). The sum of the forces elicited by the stimulation of the intact (sprouted) and cut (regenerated) roots is shown by the open squares. The contribution of the regenerating nerve fibres is illustrated by the length of the vertical line that connects the relative force of the sprouted nerve fibres (filled squares) to the force developed by both the sprouted and regenerated nerve fibres (Fig. 3B).
Nerve fibre number and size
We made more conventional measurements of nerve fibre diameters in MG nerves in the partially denervated hindlimbs as a second method to assess the effects of enlarged targets. As shown in the example of a MG nerve in Fig. 4B, 4 months after 80 % of the innervation was removed by cutting the L7 spinal root, there were obviously lower nerve fibre profiles compared with the contralateral normal MG nerve (Fig. 4A). However, despite the lower number and a trend for the nerve fibre profiles to appear more oblong in the example shown, the shapes of the fibres were generally not very different.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 4. Remaining MG nerves are not enlarged after partial denervation by spinal root section Cross-sections of MG nerves innervating normal (A) and partially denervated MG muscles 4 months (B) and 9 months (C and D) after section of the S1 spinal root. MG muscles were partially denervated by ~80 , 50 and 80 % in B, C and D, respectively. Note that the myelinated fibres in B look normal with respect to myelination except that there are more smaller fibres after partial denervation, and the axon profiles appeared to be more oblong. C, large numbers of small myelinated regenerating fibres were seen after moderate partial denervation (50 % of nerve fibres lost after S1 root cut). D, after extensive partial denervation (> 80 %), regenerated myelinated nerve fibres 'filled the gaps' and were not readily distinguished from intact nerve fibres which had sprouted to innervate partially denervated muscles. The magnification is
| ||
Nerve fibre diameters are bimodally distributed in the normal nerve because the peripheral nerve contains both sensory and motor nerve fibres belonging to types I and II afferents and and -motoneurons. The distributions of the myelinated nerve fibre diameters in MG nerves supplying a normal MG muscle and a partially denervated MG muscle, in which 80 % of the motor innervation was removed by cutting the S1 spinal root 3.5 months previously, are shown in Fig. 5. The distribution of the myelinated fibre diameters was not changed in the partially denervated muscle, consistent with the results of the charge measurements. Overall, there was no trend for the myelinated nerve fibres to increase in size and the fibre size distribution was not significantly different from normal (P > 0.05). An increase in numbers of small nerve fibres with diameters of < 3 may be indicative of early signs of axonal regeneration. At a regeneration rate of 3 mm day-1, some axonal regeneration may be expected in the 3.5 months over the ~150 mm distance from the roots to the MG nerve. Absence of any evoked muscle contractions indicated that the regenerating motor axons had not yet made functional connections.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 5. Unchanged bimodal distribution of myelinated nerve fibre diameters 3.5 months after partial denervation and putative axonal sprouting to supply enlarged target fields Bimodal distributions of the myelinated fibre diameter in nerves to MG muscle for normal (A) and partially denervated hindlimbs (B) in which 80 % of the motor innervation was removed after section of the S1 spinal root. The plots represent outer fibre diameters calculated from measurements of perimeters as described in Methods. The
| ||
Under conditions in which nerve fibres in the cut spinal root regenerated we found a significant increase in the number of small myelinated nerve fibres, particularly for longer periods of partial denervation (> 4 months; Fig. 6A and C). Examples are shown for moderate and extensive partial denervation, 50 % (Fig. 4C) and 83 % (Fig. 4D), respectively, conditions in which regenerating axons do not and do remake functional nerve-muscle connections, respectively (Rafuse et al. 1992). The many small nerve fibre profiles seen in the example of moderate partial denervation in Fig. 4C appear to 'fill' the spaces left vacated by the degenerated axons from the cut spinal root suggesting that axons in the cut spinal root had regenerated as far as the MG nerve (Fig. 4B). These nerve fibre profiles were small which is typical of regenerating nerve fibres which fail to remake peripheral connections (Gillespie & Stein,1983; Gordon et al. 1991) and consistent with our evidence that stimulation of these nerves did not evoke muscle contractions. In contrast, the nerve fibre profile in the MG nerve looked far more normal in the extensively denervated MG nerve (Fig. 4D). In this animal, the cut nerves were allowed to regenerate for 9 months which was sufficient time to permit the formation of functional nerve-muscle reconnections and allowed for good recovery of nerve properties (Gordon & Stein,1982a,b).
Frequency histograms (Fig. 6A and B) and cumulative percentages (Fig. 6C and D) of myelinated nerve fibre diameters, measured 9 months after moderate (50 %) and extensive (83 %) partial denervation, are quantified and compared in Fig. 6. When the extent of partial denervation was moderate (Fig. 6A and C), regenerating nerve fibres did not make functional contacts and consequently there was an increase in the number of small diameter myelinated nerve fibres (Fig. 6A). This can be more clearly seen when the same data are plotted as a cumulative percentage (Fig. 6C). The smaller diameters of the regenerating nerve fibres are evident from the significant leftward shift of the cumulative percentage histogram (Kolmogorov-Smirnov test, P < 0.05). When the partial denervation was extensive (83 %), many of the cut nerve fibres regenerated and re-established functional contacts with muscle. Consequently, both the frequency histogram (Fig. 6B; filled bars) and cumulative percentage histograms (Fig. 6D; filled circles) were more similar to normal. Nonetheless, the nerve fibre diameters were still significantly smaller than normal (P < 0.05) indicating that some, but not all, the regenerating nerves had made functional connections and recovered size after reinnervation (Gordon & Stein, 1982a,b). The low charge detected on the cut ventral roots (Fig. 2B) indicated that relatively small numbers of the regenerated motor nerve fibres were successful in making functional nerve-muscle connections with the partially denervated muscles.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 6. The shift of the bimodal distribution of myelinated nerve fibre diameters to smaller diameters 9 months after partial denervation is accounted for by regeneration of cut spinal roots Frequency histograms (A and B) and cumulative percentages (C and D) comparing the bimodal distributions of myelinated fibre diameters in MG motor nerves which supplied partially denervated MG muscles (filled bars;
| ||
Unitary action potentials and conduction velocity
As an independent electrophysiological measure of the size of MG motor nerve fibres which innervate partially denervated muscles, we teased ventral rootlets to evoke all-or-none unitary action potentials on the MG nerve synchronously with all-or-none and twitch-force responses in the MG muscle, as described in Methods. The latency of the unitary motor nerve action potentials was measured to calculate conduction velocity, which is positively correlated with myelinated nerve fibre diameter. We also measured the peak-to-peak amplitude which reflects fibre area (Gordon & Stein,1982b). The two electrophysiological parameters co-vary in nerve fibres of -motoneurons that supply normal (Fig. 7A) and enlarged muscles targets (Fig. 7B). Comparison of histograms for conduction velocities and unitary action potential amplitudes recorded from motor nerve fibres that innervated enlarged muscle targets after moderate (Fig. 7B and E) and extensive (Fig. 8C and F) partial denervation showed that partial denervation did not produce a significant change in either parameters in the -motoneuron population which had sprouted to reinnervate denervated muscle fibres. Although there was some tendency toward lower conduction velocities (Fig. 7B and Fig. 8B), the mean values were not significantly different (Students t test) nor were the distributions significantly different (Kolmogorov-Smirnov test) from normal.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 7. Normal relationships between action potential amplitude and conduction velocities are retained in motor nerve fibres which sprout to form enlarged motor units 6.4 ± 0.9 months after partial denervation The relationships between peak-to-peak amplitude of neural action potential and conduction velocity for single MG motor nerve fibres in normal (A) and partially denervated (B) hindlimbs are similar. The slope of the regression lines are 0.42 ± 0.05 (r = 0.75) and 0.32 ± 0.06 (r = 0.65) in A and B , respectively.
| ||
 |
View larger version [in this window] [in a new window] |
|
|
Figure 8. Unchanged mean values and frequency distributions of conduction velocity and action potential amplitude of motor nerve fibres which sprout to form enlarged motor units 6.4 ± 0.9 months after partial denervation Distributions of conduction velocities (A-C) and peak-to-peak amplitude (D-F) of unitary MG nerve action potentials recorded on the MG motor nerves in hindlimbs in which the number of motor units in MG muscles was normal (A and D), 50 % of normal (B and E) and 20 % of normal (C and F). Mean ± S.E.M values were 93.9 ± 1.01 (A), 91.5 ± 1.31 (B), 93.5 ± 1.33 (C), 23.75 ± 0.51 (D), 24.1 ± 0.54 (E) and 26 ± 0.83 (F). Neither the mean nor the distributions were significantly different (Student's t test and Kolmogorov-Smirnov test).
| ||
The mean values of the two neural parameters are plotted as a function of the extent of partial denervation (Fig. 9A and B) and as a function of tetanic force (Fig. 9C and D; an index of target enlargement). In all instances the slope of the regression line was not significantly different from zero indicating that there was no increase in either neural parameter as a function of partial denervation or target size. Thus, even within a single neuronal population, an increase in peripheral target size does not have a retrograde affect on the size of the innervating nerve fibre.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 9. Electrophysiological parameters of myelinated nerve fibre size, action potential amplitude and conduction velocity remain unchanged 6.4 ± 0.9 months after partial denervation and increase in motor unit size Motor nerve size measured electrophysiologically as means (± S.E.M.) peak-to-peak neural action potential amplitude (A and C) and conduction velocity (B and D) as a function of the relative number of remaining motoneurons (A and B) and the size of the target in terms of motor unit tetanic force (C and D). The slopes of the regression lines were not significantly different from zero.
| ||
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Increased target size does not change the size of myelinated nerve fibres
In this study we used several different and independent quantitative methods to show that an enlarged peripheral target size does not affect the size of motor or sensory nerve fibres. Thus, our findings do not support the hypothesis that target size has dynamic retrograde control over the size of myelinated nerve fibres.
Previous studies of partial denervation noted little change in conduction velocity of motor nerves with enlarged muscle targets (Brown & Ironton, 1978; Luff et al. 1988; Rafuse et al. 1992). However, sample sizes were generally small and data were pooled for partial denervations that removed as little as 10 % and as much as 90 % of the innervation. A small but significant increase in nerve fibre diameter was noted after partial denervation in an early study by Edds (1950a), but the degree of partial denervation was not determined. Our approach was to use several different quantitative electrophysiological methods to determine the extent of partial denervation, target enlargement and size of nerve fibres. All measurements concurred to show that myelinated sensory and motor nerve fibres did not increase in size in response to enlarged targets, irrespective of the extent of partial denervation and increase in target size. Taken together with the finding that even minimal contact of regenerating axons with muscle targets after botulinum toxin (BoTX) blockade was sufficient to restore normal conduction velocities (Pinter et al. 1991), the present results indicate that size of nerve fibres depends on target contact but not on the size of the target, namely the number of sense organs or muscle fibres per nerve fibre.
It is unlikely that our measurements failed to detect enlargement of nerve fibres after partial denervation because the measurements were made distal to the somatofugal progression of slow axonal transport of cytoskeletal proteins at 1-3 mm day-1. The correspondence of the proximo-distal progression of reduction and recovery of the size of nerve fibres after axotomy and reformation of target connections, respectively, with the rate of anterograde transport of neurofilament protein provided evidence for a direct relationship between axon size and the slow transport of neurofilaments (Hoffman et al. 1984, 1985). Hence, it is possible that potential changes in the size of nerve fibres which sprout after partial denervation may not have been detected by measurements made at the level of the MG nerve branch or by charge measurements made on the dorsal and ventral roots in response to stimulation of the MG nerve. However, because all measurements were made at least 3 months after partial denervation and as late as 9-12 months (Figs 4-6), any change in size of nerve fibres consequent to changes in the amounts of cytoskeletal proteins transported somatofugally at the rate of 1-3 mm day-1, should certainly have been detected in the MG nerves given that any change in fibre size would have occurred at the distance of 150 to 160 mm from the neuronal bodies within 50-60 days of somatofugal transport. The time period of 3 months was sufficient for the proteins to reach the measurement site and hence for any change in fibre size to have been detected. At longer intervals of 10 months, we observed many nerve fibres which were smaller rather than larger, these fibres being attributed to axonal regeneration from the cut root (Fig. 5). At this time point, we certainly would have seen larger than normal nerve fibres had there been transport of increased numbers of neurofilaments consequent to the enlarged target innervation. Hence, the inability to detect enlargement of nerve fibre size in association with enlarged targets, using both electrophysiological and morphological methods, is consistent with the conclusion that enlarged targets do not have a retrograde influence on the size of myelinated nerve fibres which remains unchanged despite an enlargement of their peripheral target.
The possibility exists that stable charge values for dorsal and ventral root nerves mask an increase in size and charge of some fibres and a decline in others which, in turn, could cancel each other to produce no change in the integrated values. However, this possibility is unlikely in view of the insignificant changes in action potential amplitude and conduction velocities of single motor nerve fibres which supply enlarged muscle units (Fig. 8 and Fig. 9). Moreover, measurement of nerve fibre diameters never revealed enlarged fibre sizes (Figs 4-6).
Our findings that enlarged targets failed to increase myelinated nerve fibre size contrasts with findings that: (1) enlarged targets by partial denervation dramatically increased the size of non-myelinated sympathetic nerve fibres (Voyvodic, 1989) and (2) high secretory activity of tractus hypophyseus axons is associated with an increase in axonal diameter (Grainger & Sloper, 1976). The latter findings for unmyelinated nerve fibres support the hypothesis that target size controls nerve fibre calibre by retrograde uptake of neurotrophins (Voyvodic, 1989) in a fashion analogous to the target regulation of neuronal survival demonstrated in classical embryological studies in which target size was increased or decreased (Korsching, 1993). Findings that target levels of nerve growth factor generally reflect the density of innervation of sympathetic target cells are also consistent with the original hypothesis (Korsching & Thoenen, 1983; Shelton & Reichardt, 1984). The lack of adaptation of myelinated fibre size to target size demonstrated here shows that myelinated and non-myelinated nerve fibres differ with respect to adult plasticity, at least so far as the calibre of their fibres.
The simplest explanation for the contrasting abilities of non-myelinated and myelinated axons to increase in size lies in the absence and presence of a myelin sheath, respectively. In their study of the atrophy of myelinated nerve fibres after axotomy, Gillespie & Stein (1983) made the important observation that, while axonal area declined after axotomy, myelin thickness remained unchanged. As a result, the axon literally collapsed within the confines of the myelin sheath. The corollary was that restoration of normal fibre size after reformation of target connections (Gordon & Stein, 1982a) occurred by expansion of the axonal size. If this explanation is correct, and the target size regulates the size of the axons, it follows that axons can increase to a maximum of the confines of the myelin sheath. That maximum would be reached when the nerve fibres are a full circle, the minimal circularity being the limit of atrophy (Gillespie & Stein, 1983) and the maximal circularity being the limit of hypertrophy. Although the present study cannot conclusively confirm this simple explanation, nerve fibres in the cross-sections taken from extensively partially denervated MG muscles do not appear to have reached their maximum circularity (compare Fig. 4A and B). Consequently, the nerve fibres may not have reached their maximum size even if the myelin sheath ultimately acts to restrain axon calibre enlargement. However, it should be noted that the closest apposition between the Schwann cell and axon occurs at the paranodal region, and not the internodal region measured in this study. Thus, we cannot completely rule out the possibility that the paranodal region ultimately determines the maximal axon size in the adult.
Activity-mediated rather than target-size regulation of nerve fibre size
The most dramatic changes in myelinated nerve fibre size occur after axotomy where reduced conduction velocity and extracellular action potential amplitude are part of a repertoire of changes in electrophysiological properties of axotomised motoneurons (Gufstaffson & Pinter, 1984; Foehring et al. 1986b; Titmus & Faber, 1990). As axonal size of axotomized neurons declines without any change in myelin thickness (Gillespie & Stein, 1983), reduced nerve fibre size can be attributed, at least in part, to the downregulation in the expression of neurofilament proteins in the axotomized neurons (Hoffman et al. 1984; Tetzlaff et al. 1988). Our findings here, and those of Pinter et al. (1991), indicate that it is the reformation of target connections and not the size of the target that is important in reversing the axotomy-induced downregulation of neurofilament mRNA, and in turn, promoting the recovery of the size of the reinnervating nerve fibres. Evidence that direct application of the neurotrophic factors, BDNF, NT-3 or NT-4/-5, to the stump of axotomised motoneurons prevents the normal reduction in conduction velocity of the axotomised motoneurons (Munson et al. 1997b), strongly indicates that the neurotrophic factors from the target play a key role in determining size of peripheral nerves (cf. Gold et al. 1991). However, our findings indicate that it is access to neurotrophic factors provided by the target, rather than the size of the target that regulates the size of the nerve fibres.
There is now experimental evidence that: (1) the production and uptake of neurotrophic factors is an activity-dependent process rather than a process dependent on the size of the peripheral target and (2) this activity-dependent production of neurotrophic factors may regulate both the size and electrophysiological properties of motoneurons. Production of NT-4/-5 in rat skeletal muscle depends on muscle activity with mRNA synthesis declining after blockade of neuromuscular transmission and increasing with imposed nerve stimulation in the adult, especially in slow muscle fibres (Funokoshi et al. 1995). We recently reported that, in association with fast to slow conversion of muscle units by low frequency chronic electrical stimulation, the mean conduction velocity of the stimulated motoneurons declined significantly (Gordon et al. 1997) concurrent with some conversion of other motoneuron properties towards the S-type (Munson et al. 1997a). The stimulation-induced decline in conduction velocity is consistent with the activity-dependent production of NT-4/-5 by the stimulated muscle fibres (Funokoshi et al. 1995) and can explain earlier observations of reductions in the size of hindlimb motoneuron cell bodies and their myelinated nerve fibres after intense exercise regimes (Anderson & Edstrom, 1957; Roy et al. 1983).
In summary, our experimental results suggest that myelinated nerve fibre size depends on contact with target and is not dependent on the number of target connections. In contrast, the fibre size can be altered by activity within the adult range (Roy et al. 1983; Gordon et al. 1997; Munson et al. 1997a), possibly associated with the uptake of neurotrophic factors from the active target (Mendell et al. 1994). Hence it is likely that it is the activity of the nerves, and not the number of target connections, that modulates their size, thus demonstrating a dynamic regulation of size of myelinated nerve fibres by activity rather than simply by the number of target connections.
Motoneuronal somatic plasticity under conditions which promote growth
In contrast to nerve fibre size, soma size of motoneurons increases significantly after partial denervation and putative sprouting (Tissenbaum & Parry, 1992). The soma enlargement was attributed to the increased number of reinnervated muscle fibres because the increase was prevented when some muscle tissue was extirpated at the same time as the partial denervation. Electrophysiological properties of motoneurons other than conduction velocity are also sensitive to both the number and size of muscle fibres innervated by the motoneurons. The duration of the afterhyperpolarization (AHP) in soleus motoneurons declined in response to motor unit enlargement after partial denervation (Huisar et al. 1977) as well as to a reduction in the size of the muscle fibres produced by TTX-induced nerve conduction block, BoTX-induced muscle paralysis, spinal cord transection and limb immobilization (Czeh et al. 1978; Gallego et al. 1979; Cope et al. 1986; Pinter & van den Noven, 1991). Interestingly, all the experimental conditions induce a 'growing' or 'regenerative' state in the affected neurones.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ANDERSSON Y. & EDSTROM, J. (1957). Motor hyperactivity resulting in diameter decrease of peripheral nerves. Acta Physiologica Scandanavica 39, 240-245 | |
| BOYD I. A. & DAVEY, M. R. (1966). The composition of peripheral nerves. In Control and Innervation of Skeletal Muscle, ed. ANDREW B.L. pp 35-47. Churchill Livingstone, Edinburgh | |
| BROWN M. C. & IRONTON, R. (1978). Sprouting and regression of neuromuscular synapses in partially denervated mammalian muscles. Journal of Physiology 278, 325-348. | [Abstract] |
| COPE T. C., BODINE, S. C., FOURNIER, M. & EDGERTON, V. R. (1986). Soleus motor units in chronic spinal transected cats: physiological and morphological alterations. Journal of Neurophysiology 55, 1202-1220 | [Medline] |
| CZEH G., GALLEGO, R., KUDO, N. & KUNO, M. (1978). Evidence for the maintenance of motoneurone properties by muscle activity. Journal of Physiology 281, 239-252. | [Abstract] |
| EDDS M. V. (1949). Experiments on partially deneurotized nerves. Journal of Experimental Zoology 111, 211-226 | |
| EDDS M. V. (1950a). Collateral regeneration of residual motor axons in partially denervated muscles. Journal of Experimental Zoology 113, 517-552. | |
| EDDS M. V. (1950b). Hypertrophy of nerve fibers to functionally overloaded muscles. Journal of Comparitive Neurology 93, 259-275. | |
| FISZ M. (1963). Probability and Mathematical Statistics. Wiley, New York. | |
| FOEHRING R. C., SYPERT, G. W. & MUNSON, J. B. (1986a). Properties of self-reinnervated motor units of medial gastrocnemius of cat. I. Long-term reinnervation. Journal Neurophysiology 57, 931-946 | |
| FOEHRING R. C., SYPERT, G. W. & MUNSON, J. B. (1986b). Properties of self-reinnervated motor units of medial gastrocnemius of cat. II. Axotomized motoneurons and time course of recovery. Journal of Neurophysiology 57, 947-965 | |
| FU S. & GORDON, T. (1997). The cellular and molecular basis of peripheral nerve regeneration. Molecular Neurobiology 14, 67-116 | [Medline] |
| FUNOKOSHI H., BULLUARDO, N., ARENAS, E., YAMMAMOTO, Y., CASABONA, A., PERSSON, H. & IBANEZ, C. F. (1995). Muscle-derived neurotrophin-4 as an activity-dependent trophic signal for adult motor neurons. Science 268, 1495-1499. | [Medline] |
| GALLEGO R., KUNO, M., NUNEZ, R. & SNIDER, W. D. (1979). Dependence of motoneurone properties on the length of immobilized muscle. Journal of Physiology 291, 179-189 | [Medline] |
| GILLESPIE M. J. & STEIN, R. B. (1983). The relationship between axon diameter, myelin thickness and conduction velocity during atrophy in mammalian peripheral nerves. Brain Research 259, 41-56. | [Medline] |
| GOLD B. G, MOBLEY, W. C. & MATHESON, S. F. (1991). Regulation of axonal caliber, neurofilament content, and nuclear localzation in mature sensory neurons by nerve growth factor. Journal of Neuroscience 11. 943-955 | |
| GORDON T. (1983). Dependence of peripheral nerves on their target organs. In Somatic and Autonomic Nerve-Muscle Interactions, ed. BURNSTOCK, G., VRBOVA, G. & O'BRIEN, R. A. D., pp. 289-325. Elsevier, | |
| GORDON T., GILLESPIE, J., OROZCO, R. & DAVIS, L. (1991). Axotomy-induced changes in rabbit hindlimb nerves and the effects of chronic electrical stimulation. Journal of Neuroscience 11, 2157-2169 | [Abstract] |
| GORDON T. & STEIN, R. B. (1982a). Time course and extent of recovery in reinnervated motor units of cat triceps surae muscles. Journal of Physiology 323, 307-323. | [Medline] |
| GORDON T. & STEIN, R. B. (1982b). Reorganization of motor-unit properties in reinnervated muscles of the cat. Journal of Neurophysiology 48, 1175-1190. | [Medline] |
| GORDON T., TYREMAN, N., RAFUSE, V. F. & MUNSON, J. B. (1997). Fast-to-slow conversion following chronic low-frequency activation of medial gastrocnemius muscles in cats. I. Muscle and motor unit properties. Journal of Neurophysiology 77, 2585-2604 | [Abstract/Full Text] |
| GRAINGER F. & SLOPER, J. C. (1976). Microtubular number in the tractus hypophyseus of newborn normal rats and newborn rats with congenital diabetes insipidus. Cell and Tissue Research 169, 405-414 | [Medline] |
| GUFSTAFSSON B. & PINTER, M. (1984). Effects of axotomy on the distribution of passive electrical properties of cat motoneurones. Journal of Physiology 356, 433-442 | [Abstract] |
| HOFFER J. A., STEIN, R. B. & GORDON, T. (1979). Differentiation atrophy of sensory and motor fibers following section of cat peripheral nerves. Brain Research 178, 347-361. | [Medline] |
| HOFFMAN P. N., CLEVELAND, D. W., GRIFFEN, J. W., LANDES, P. W., COWAN, N. J. & PRICE, D. L. (1987). Neurofilament gene expression: a major determinant of axon caliber. Proceedings of the National Academy of Sciences of the USA 84, 3472-3476 | [Medline] |
| HOFFMAN P. N., GRIFFEN, J. W. & PRICE, D. L. (1984). Control of axonal caliber by neurofilament transport. Journal of Cellular Biology 99, 705-714. | |
| HOFFMAN P. N., LASEK, R. J., GRIFFIN, J. W. & PRICE, D. L. (1983). Slowing of the axonal transport of neurofilament proteins during development. Journal of Neuroscience 3, 1694-1700 | [Abstract] |
| HOFFMAN P. N., THOMPSON, G. W., GRIFFEN, J. W. & PRICE, D. L. (1985). Changes in neurofilament transport coincide temporally with alterations in the caliber of axons in regenerating motor fibers. Journal of Cellular Biology 101, 1332-1340. | |
| HUISAR P., KUDO, N., KUNO, M. & MIYATA, Y. (1977). Reaction of intact spinal motoneurones to partial denervation of the muscle. Journal of Physiology 265, 175-191. | [Abstract] |
| KORSCHLING S. (1993). The neurotrophic factor concept: a reexamination. Journal of Neuroscience 13, 2739-2748 | [Abstract] |
| KORSCHLING S. & THOENEN, H. (1983). Nerve growth factor in sympathetic ganglia and corresponding target organs of the rat: correlation with density of sympathetic innervation. Proceedings of the National Academy of Sciences of the USA 80, 3513-3516. | [Medline] |
| KUNO M., MIYATA, Y. & MUNOZ-MARTINEZ, E. J. (1974a). Properties of fast and slow -motoneurones following motor reinnervation. Journal of Physiology 242, 273-288 |
[Medline] |
| KUNO M., MIYATA, Y. & MUNOZ-MARTINEZ, E. J. (1974b). Differential reactions of fast and slow -motoneurones to axotomy. Journal of Physiology 240, 725-739. |
[Medline] |
| LUFF A. R., HATCHER, D. D. & TORKKO, K. (1988). Enlarged motor units resulting from partial denervation of cat hindlimb muscles. Journal of Neurophysiology 59, 1377-1394 | [Medline] |
| MENDELL L. M., COLLINS, W. F. III & MUNSON, J. B. (1994). Retrograde determination of motoneuron properties and their synaptic input. Journal of Neurobiology 25, 707-721 | [Medline] |
| MUNSON J. B., FOEHRING, R. C., MENDELL, L. M. & GORDON, T. (1997a). Fast-to-slow conversion following chronic low-frequency activation of medial gastrocnemius muscles in cats. I. Motoneuron properties. Journal of Neurophysiology 77, 2605-2615 | [Abstract/Full Text] |
| MUNSON J. B., SHELTON, D. L. & MCMAHON, S. B. (1997b). Adult mammalian sensory and motor neurons: roles of endogenous neurotrophins and rescue by exogenous neurotrophins after axotomy. Journal of Neuroscience 17, 470-476 | [Abstract/Full Text] |
| PINTER M. J., VAN DEN NOVEN, S., MUCCIO, D. & WALLACE, N. L. (1991). Axotomy-like changes in cat motoneuron electrical properties elicited by botulinum toxin depend on complete elimination of neuromuscular transmission. Journal of Neuroscience 11, 657-666. | [Abstract] |
| RAFUSE V. F. & GORDON, T. (1996). Self-reinnervated cat medial gastrocnemius muscles. I. Comparison of the capacity for regenerating nerves to form enlarged motor units after extensive peripheral nerve injuries. Journal of Neurophysiology 75, 268-281 | [Medline] |
| RAFUSE V. F., GORDON, T. & OROZCO, R. (1992). Proportional enlargement of motor units following partial denervation of cat triceps surae muscles. Journal of Neurophysiology 68, 1261-1276 | [Medline] |
| ROY R. R., GILLIAM, T. B., TAYLOR, J. F. & HEUSNER, W. W. (1983). Activity-induced morphologic changes in rat soleus nerve. Experimental Neurology 80, 622-632 | [Medline] |
SHELTON D. L. & REICHARDT, L. F. (1984). Expression of  -nerve growth factor gene correlates with density of sympathetic innervation in effector organs. Proceedings of the National Academy of Sciences of the USA 81, 7951-7955 |
[Medline] |
| TETZLAFF W., BISBY, M. A. & KREUTZBERG, G. W. (1988). Changes in cytoskeletal proteins in the rat facial nucleus following axotomy. Journal of Neuroscience 9, 914-922. | [Abstract] |
| TISSENBAUM H. A. & PARRY, D. J. (1992). The effect of partial denervation of tibialis anterior muscle on the number and sizes of motoneurons in TA motor nucleus of normal and dystrophic (C57BL dy2j/dy2j) mice. Canadian Journal of Physiology and Pharmacology 69, 1769-1773 | |
| TITMUS M. J. & FABER, D. S. (1990). Axotomy-induced alterations in electrophysiological characteristics of neurons. Progress in Neurobiology 35, 1-51. | [Medline] |
| VEJSADA R., PALECEK, J., HNIK, P. & SOUKUP, T. (1985). Postnatal development of conduction velocity and fibre size in the rat tibial nerve. International Journal of Developmental Neuroscience 3, 583-595 | |
| VOYVODIC J. I. (1989). Target size regulates calibre and myelination of sympathetic axons. Nature 342, |
Acknowledgements
This work was supported by the Muscular Dystrophy Association and Medical Research Council of Canada. We thank Drs Roberto Orozco and Jean Gillespie for their contribution to the early part of this work.
Corresponding author
Tessa Gordon: Division of Neuroscience, 525, Heritage Medical Research Centre, Faculty of Medicine, University of Alberta, Edmonton, Canada T6G 2S2.
Email: tessa.gordon{at}ualberta.ca
Author's present address
V. F. Rafuse: Department of Anatomy and Neurobiology, Sir Charles Tupper building, Dalhousie University, Halifax, Canada B3H 4H7.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and ventral (
) roots and B the tetanic force of MG muscles in response to stimulation of the uncut ventral roots (
) in response to supramaximal stimulation of the MG nerve, 6.4 ± 0.9 months after partial denervation. Values are expressed relative to charge delivered to the two contributing roots (A) and force elicited by the both ventral roots (B) on the contralateral control side. Both are plotted as a function of the proportion of axons lost in the cut root (percent partial denervation). The data points are compared with a linear decline of relative force and charge (diagonal lines) which would be predicted if there was no compensatory change in nerve fibre size (A) and the numbers of muscle fibres supplied by each motoneuron (B). Further details in the text.
) with those supplying contralateral control muscles (open bars; 
) 9 months after removing either 50 % (A and C) or 83 % of the motor innervation (B and D). There were many more small myelinated fibres in the MG nerves when the muscle was partially denervated by 50 % compared with 83 % (i.e. compare C and D). Nonetheless, the distributions were significantly different in both cases (Kolmogorov-Smirnov test, P < 0.05). The more substantial difference between the size of the myelinated nerve fibre diameters after 50 % as compared with 83 % partial denervation could be accounted for by regeneration of cut axons from the cut spinal roots which resulted in functional reinnervation of MG muscle fibres only after extensive denervation (83 %, B and D).