- Email: [email protected]
Neonatal paralysis of the rat soleus muscle selectively affects motoneurones from more caudal segments of the spinal cord
DEVELOPMENTAL BRAIN RESEARCH
ELSEVIER
DevelopmentalBrain Research 98 (1997) 281-286
Research report
Neonatal paralysis of the rat soleus muscle selectively affects motoneurones from more caudal segments of the spinal cord Linda Greensmith, Angela Hind, Gerta Vrbovfi * Department of Anatomy and Developmental Biology, Centre for Neuroscience, University College London, Gower Street, London, WCIE 6BT, UK Accepted 29 October 1996
Abstract Transient paralysis of the rat soleus muscle shortly after birth leads to permanent muscle weakness, loss of muscle fibres and the death of motoneurones. The soleus muscle receives its innervation from motoneurones whose axons exit the spinal cord either via the L4 ventral ramus or in the more caudal part of the cord via the L5 ventral ramus. Whether both populations of motoneurones are equally affected by neonatal paralysis was studied here. In soleus muscles paralysed with a-bungarotoxin shortly after birth and examined 8-10 weeks later, there is no loss of force or muscle fibres in the part of the muscle supplied by axons in the L4 ventral ramus. Loss of force and muscle fibre numbers occurs only in the part of the muscle supplied by axons in the L5 ventral ramus. In a normal adult soleus 30.3 + 2.4% of muscle force is produced by stimulating the L4 ventral ramus and 69.0 ___5.5% by stimulating the L5 ventral ramus. In soleus muscles treated with a-bungarotoxin 28 _ 1.4% of the force produced by the contralateral control soleus was generated by axons in the L4 ventral ramus and only 20.3 + 5.6% by stimulating the L5 spinal nerve. The number of muscle fibres supplied by either ventral ramus in control and experimental muscles confirmed that the decrease of force after treatment with a-bungarotoxin can be accounted for by loss of muscle fibres supplied by axons in the L5 ventral ramus. The reduced force production and muscle fibre numbers was due to a selective loss of motoneurones that had their axons in the L5 ventral ramus. The number of axons to soleus in the L4 ventral ramus was 9.3 + 0.7 in controls and 10.3 _ 0.9 in the experimental animals, whereas the L5 ventral ramus contained 17.2 + 0.7 in controls and only 4.7 + 1.7 in the experimental animals. Thus paralysis of the soleus muscle at birth selectively affects motoneurones in the more caudal part of the spinal cord, suggesting that the more cranial motoneurones are more mature and less likely to be influenced by lack of neuromuscular interaction at the time of birth. Keywords: Motoneurone;Nerve:muscleinteraction;Development;Paralysis
1. Introduction It is well established that during the early postnatal period motoneurones remain dependant upon contact with their target muscle for l~heir survival. Following nerve injury at birth, more than 70% of motoneurones to the soleus muscle die [5,11,18]. This death of target-deprived motoneurones is a consequence of the lack of functional interaction between the motoneurone and its target muscle, since disruption of nerve:muscle interaction during this critical period of development also results in motoneurone death. We have previously reported that a brief, transient paralysis of the soleus muscle in neonatal rats results in the death of a large proportion of motoneurones to the soleus
* Corresponding author. Fax: +44 (171) 380 7349. E-mail: [email protected]
muscle [7]. Moreover, the soleus muscles paralysed during this critical period of postnatal development remained permanently weak, due to the loss of a proportion of their muscle fibres [4]. The extent of the reduction of force and loss of muscle fibres correlated with the decrease in motoneurone numbers. When the muscle was paralysed for only a short period of time, around 30% of motoneurones to the soleus muscle died and there was an equivalent reduction in muscle force. Prolonging the duration of the paralysis resulted in a greater loss of both motoneurones and muscle fibres. Thus it appears that following disruption of neuromuscular interaction shortly after birth, the main reason for the consequent muscle weakness was the loss of motoneurones. However, in these experiments, following neonatal muscle paralysis, not all the motoneurones to the soleus muscle died, so that some of the motoneurones possessed properties that rendered them resistant to the effects of target deprivation. Therefore there
0165-3806/97/$17.00 Copyright © 1997 Elsevier Science B.V. All rights reserved. PII SO 165 -3 806(96)00200-3
282
L. Greensmithet al./ DevelopmentalBrain Research 98 (1997) 281-286
are clearly differences in the responses of motoneurones to disruption of neuromuscular interaction. Neuromuscular interaction in neonatal rats can be disrupted by blocking the postsynaptic acetylcholine receptor (ACh) with the snake toxin, c~-Bungarotoxin (BTX). This treatment of the neuromuscular junction of the soleus muscle in newborn rats, results in muscle paralysis that lasts for 2 - 3 days, and this paralysis can be prolonged by additional treatments with BTX [7]. In this study we wished to establish which motoneurones are more likely to be affected by transient neonatal paralysis. The soleus muscle receives its innervation via two ventral roots: 30% of motor axons to soleus exit through the L4 spinal nerve (SN) and these are axons from motoneurones situated in the more cranial segment of the cord; 70% of motor axons are in the L5 SN and most of these are from the more caudal part of the soleus motor pool. In view of the cranio-caudal progression of the maturation of the spinal cord, motoneurones in the L4 segment might be expected to be more mature than those in the L5 part of the soleus motoneurone pool. It is well known that the susceptibility of motoneurones to separation from their target declines rapidly with age [18] (for reviews see [8,13]). For example, we have previously found that sciatic nerve crush in newborn rats results in the death of as many as 90% of motoneurones [5], injury at 3 days causes the loss of up to 80% [2], but the same insult carded out in 5-day-old rats causes little if any motoneurone death [3,10]. It is therefore possible that the more cranial, mature motoneurones in the L4 segment are less likely to be affected by temporary loss of interaction with the muscle fibres they supply. The present study was designed to examine this possibility.
silicon rubber (Dow Coming, 3140 RTV), weighing 0.3 mg and containing approximately 7 / z g BTX and 120 /zg NaC1 were implanted into the hindlimb of newborn rats. Strips weighing 0.75 mg and containing 16 /xg BTX and 240 /~g NaC1 were implanted into the hindlimbs of 3 day old rats. Control implants of similar weights were implanted into control animals and contained 120 /zg and 240 /xg NaC1 at each of the respective ages. 2.3. Isometric tension recordings When the rats were 2 - 1 0 months old, they were anaesthetised with chloral hydrate (4%, 1 m l / 1 0 0 g body weight, i.p.) and the soleus muscles in both the treated and contralateral control legs were prepared for in vivo assessment of muscle force. The soleus muscle in each leg was dissected free of its surrounding tissue. Both legs were rigidly secured to a table and the distal tendons attached to isometric force transducers (Dynameter UFI, Devices) via silk threads. The sciatic nerve was exposed in the leg, isolated from the surrounding tissue and followed proximally to the point where the L4 and L5 spinal nerves (SN) join to form the sciatic. The spinal nerves were then dissected free as far proximally as possible, sectioned and ligated. In the contralateral control leg, the soleus nerve was dissected. All other muscles of the hindlimbs were then denervated and their distal tendons severed. The length of the soleus muscles in both legs was adjusted to that at which the muscles developed maximal twitch tension. Isometric contractions were elicited by stimulating the cut end of either the L4 or L5 SN, or the motor nerve to the soleus muscle, using a pulse width of 0.02 ms. To elicit tetanic contraction the muscles were stimulated at 20, 40 and 80 Hz for 800 ms.
2. Materials and methods
2.4. Motor unit numbers
2.1. Surgical procedures
To estimate the total number of motor units in the treated and contralateral control soleus muscles the motor nerve to soleus was stimulated by single pulses every 4 s. To estimate the number of motor axons in each SN to the treated soleus muscle, the distal end of each SN was individually stimulated and the twitch tension of soleus recorded. The stimulus intensity applied to the nerve was gradually increased to obtain stepwise increments of twitch tension as individual motor axons with increasing stimulus thresholds are recruited. The number of stepwise increments was counted.
Under halothane anaesthesia and sterile conditions, a small silicon rubber strip, containing either o~-bungarotoxin (BTX, Sigma) or NaC1, was implanted alongside the soleus muscle in the right hindlimb of neonatal Wistar rats, within 6 - 1 2 h of birth. The contralateral leg was left unoperated and served as a control. When the pups had fully recovered from the anaesthesia they were returned to their mothers. Some of these operated animals received a second silicon strip 3 days after the initial operation. Under the same conditions as above, the first implant was identified and removed and replaced with a larger implant containing either BTX or NaC1, as appropriate. On recovery from the anaesthesia these animals were returned to their mothers and reared as normal. 2.2. Preparation of implants The preparation of BTX containing implants has previously been described [6]. Silicon strips prepared from inert
2.5. Glycogen depletion In some of the experiments in which the maximum tetanic tension of fibres innervated by a specific SN had been estimated, the technique of glycogen depletion was employed as a means of identifying the number and location of the muscle fibres innervated by axons emanating from either of the SN. Glycogen within the muscle fibres
L. Greensmith et al. / Developmental Brain Research 98 (1997) 281-286
283
was depleted by repetitive stimulation of the muscle at 40 Hz for 250 ms, via either the L4 or the L5 SN. The muscles were usually stimulated for 4 5 - 6 0 min, or until the tetanic contraction produced in response to stimulation had decreased to approximately 5% of the original. 2.6. Muscle histology
After the tension experiments had been completed, the control and treated soleus muscles were dissected out from the rats under terminal chloral hydrate anaesthesia. The muscles were then stretched to a length similar to that in the leg and the control and treated muscle from each animal was mounted side by side and frozen in melting isopentane. To determine the number of muscle fibres, transverse cryostat sections (10 /xm) were collected onto gelatinised slides and stained for succinate dehydrogenase (SDH) [14]. Sections taken from muscles that had undergone glycogen depletion of fibres innervated by axons of either the L4 or the L5 SN were processed for the periodic acid-Schiff (PAS) reaction for glycogen [16]. This method demonstrates the presence of glycogen in muscle fibres. The number of muscle fibres innervated by each SN could then be distinguished, depending on their staining with the PAS stain. Those fibres which had been depleted of their glycogen stores by repetitive stimulation via their innervating SN did not react with the PAS stain. The number of muscle fibres in each soleus muscle was counted using a camera lucida attached to a light microscope. In those muscle in which the muscle fibres innervated by one of the SN lind been depleted of glycogen, both the total number of muscle fibres and the number innervated by each of the SN was counted.
3. Results 3.1. Tetanic tension in response to stimulation of either the L4 or L5 spinal nerve
The contribution of the: tension elicited by stimulating either the L4 or the L5 SN to the total force output (maximum tetanic tension or MTT) of the normal soleus muscle of adult rats was assessed in 7 animals aged between 2 and 10 months. An example of the tension recording from such an experiment is shown in Fig. la. Table 1 summarises the results and shows that in normal soleus muscles stimulation of the IA SN elicited 30.7% ( _ 2.7 S.E.M., n = 7) of the total muscle force, and stimulation of the L5 SN elicited 69.5% (+__7.5 S.E.M., n = 7). These results are consistent with values reported by other authors [12]. Soleus muscles that had been transiently paralysed by treatment with a single BTX implant shortly after birth were studied next. Fig. lb shows an example from an
b
25g[ 100m~ Fig. 1. Examples of isometric muscle contractions from adult soleus muscles are shown. Trace (a) is from a normal, control soleus muscle, (b) from a muscle treated with a single BTX implant at birth, and (c) is from a muscle treated with 2 BTX implants, at birth and 3 days of age. The maximum tetanic tension elicited by stimulating either the entire sciatic nerve, or the IA or L5 spinal nerve is illustrated.
experiment in which the MTT of adult soleus muscles was recorded in response to stimulation of either the L4 or L5 SN or the entire sciatic nerve. The results from the experiments in this group are summarised in Table 1, which shows that the soleus muscles treated with BTX at birth produced 40.5% ( _ 2.6 S.E.M., n = 7) of their MTT in response to stimulation of the L4 SN and 63% ( + 2 . 9 S.E.M., n = 7) in response to stimulation of the L5 SN. Examples of records taken from a soleus muscle that had been treated shortly after birth with 2 successive implants containing BTX and which were therefore paralysed for a longer period of time, are shown in Fig. lc. Table 1 summarises the results and shows that adult soleus muscles treated in this way during the early postnatal period develop 59.4% ( + 7.4 S.E.M., n = 4) of their total force in response to stimulation of their L4 SN, and 39.3% ( _ 7.4 S.E.M., n = 4) in response to stimulation of the L5 SN. Thus after treatment with BTX shortly after birth, the proportional contribution of the L4 and L5 SN to the force output of the adult soleus muscle has changed. Fig. 2 illustrates this result. Since the BTX-treated soleus muscles are permanently weaker than untreated control muscles and have fewer
L. Greensmith et al. / Developmental Brain Research 98 (1997) 281-286
284
A
90 80.
BL4 --i--
raL5
70,
Motor unit number
60.
1.,4
L5
9.3 5:0.7 10.5 + 1.0 10.3 + 0.9
17.2 + 0.7 15.0+0.9 4.7 ___1.7
50
.~ 40 "" 30, [- 20.
Control (7) 1 BTX (7) 2 BTX (4)
J::
Control (7)
NaCI (4)
1 BTX
2 BTX
(7)
(4)
The mean number of motor units ( 5: S.E.M.) from adult soleus muscles in response to stimulation of either the L4 or L5 spinal nerve is shown, in control animals and animals which received either a single BTX implant at birth, or 2 BTX implants, at birth and 3 days, and were examined 8-10 weeks later. The number of animals in each group is shown in parentheses.
B
9
Table 2 Number of motor units in adult soleus after neonatal treatment with BTX
120
IlL4
100
QL5
N
80
60
Fig. 2 summarise the results and show that the reduced force seen in the muscles treated with BTX can be entirely accounted for by the reduced contribution of the L5 SN to the total force output of the muscle.
40
3.2. Motor unit numbers
20 0 Control (7)
NaCI (4)
1 BTX
2 BTX
(7)
(4)
Fig. 2. The block diagrams illustrate the maximum tetanic tension elicited in adult animals from normal, control soleus muscles and those treated with either NaC1, a single BTX implant at birth, or 2 BTX implants at birth and 3 days of age, in response to stimulation of either the L4 or L5 spinal nerve, and expressed (A) as a % of the total muscle force, and (B) as a % of the total force of the contralateral control soleus muscle. The number of animals in each group is shown in parentheses, and the error bars represent the S.E.M..
muscle fibres (see [4]), it is possible that this alteration in the contribution to force output produced by stimulation of either SN is caused by a disproportionately greater loss of force from the part of the muscle supplied by the L5 spinal nerve. Therefore, in each animal the MTT obtained by stimulating either spinal nerve was expressed as a percentage of the total force produced by stimulating the soleus motor nerve in the contralateral control leg. Table 1 and
We have previously shown that following transient paralysis of the soleus muscle at birth, there is a loss of motoneurones [7] and a reduction in the number of surviving motor axons in the soleus nerve [4]. Whether this loss of motoneurones is evenly distributed between the L4 and L5 lumbar segments was studied next. Table 2 summarises the results of the number of motor units obtained by recording the stepwise increments of single twitch tension from the soleus muscle in response to stimulation of either the L4 or L5 SN by stimuli of increasing intensity. The table shows that the mean number of axons in the L4 SN is approximately the same in all the experimental groups, but the number of axons in the L5 SN is lower in animals treated with BTX implants at birth, and this decrease is much greater in the group of animals treated with 2 BTX implants. These results indicate that paralysis of the soleus muscle at birth affects mainly those motoneurones that send their axons through the L5 SN and that those that leave the spinal cord through the more cranial nerve are spared.
Table 1 Maximum tetanic tension of soleus muscle after neonatal treatment with BTX Treatment
Maximum tetanic tension Total muscle force (%op/con)
Control (7) NaC1 (4) 1 BTX (7) 2 BTX (4)
98.0 + 6.7 117 + 7.3 73 + 4.5 49.5 -I- 4.3
Stimulation of L4 or L5 SN (% total force)
Stimulation of L4 or L5 SN (% total force of control soleus)
L4
L4
30.7 20.0 40.5 59.4
L5 + + _ +
2.7 10.2 2.6 7.4
69.0 80.0 63.0 39.3
_ 3.5 _ 8.6 + 2.9 -I- 7.4
30.3 27.4 28.4 28.7
L5 + + + +
2.4 11 3.5 1.4
69.0 95.6 46.1 20.3
-I- 5.5 + 8.1 + 3.9 5:5.6
The mean values ( ___S.E.M.) of muscle force (MTT) developed by control soleus muscles of adult animals, and by soleus muscles that were treated with NaC1, a BTX implant at birth or 2 BTX implants, at birth and 3 days later, are expressed as a percentage of the control muscle. In addition, the MTT elicited by stimulation of either the L4 or L5 spinal nerve is shown. These values are expressed as a percentage of the MTT developed by the experimental muscle, and as a percentage of the MTT of the contralateral control soleus. The number of animals in each group is shown in parentheses.
L. Greensmith et aL/ Developmental Brain Research 98 (1997) 281-286
285
Table 3 Number of soleus muscle fibres innervatedby the IA and L5 spinal nerves Treatment
Control (7) NaC1 (4) 1 BTX (7) 2 BTX (2)
Number of musclefibres
% Of total numberof musclefibres
% Of total numberof muscle fibres in control soleus
L4
L5
L4
L5
L4
L5
812 + 125 786 + 149 753 ± 102 1705
1807 ___102 2007 + 76 1555 + 317 112
30.8 + 3.7 28.0 + 4.4 30.0 ± 7.1 94
69 + 3.2 72 + 4.4 64 _+7.1 6
20.8 + 3.6 27.3 + 3.5 30 _ 5.5 64
69.3 + 2.0 69.8 + 2.0 56.2 + 7.6 4
Glycogen was depleted from muscle fibres suppliedby axons in the LA spinal nerve by repeated stimulation.This permitted identificationof the muscle fibres innervatedby axons in either the L4 or L5 spinal nerve. The mean number of glycogendepleted muscle fibres ( + S.E.M.) innervatedby the L4 and L5 spinal nerves is shown for control soleus muscles and those treated with either NaC1, 1BTX implant at birth, or 2 BTX implants at birth and 3 days of age. The results are also expressed as a percentage of the total number of muscle fibres present, and as a percentage of the total numberof muscle fibres in the contralateralcontrol soleus muscles. The number of animals in each group is given in parentheses.
3.3. M u s c l e f i b r e n u m b e r s
The next set of experiments was carded out in order to establish the numbers of muscle fibres supplied by axons in the L4 or the L5 SN respectively in both normal soleus muscles and those treated with a single or double BTX implant. The number of muscle fibres innervated by each of the SN was investigated using the technique of glycogen depletion [9]. The muscle fibres innervated by one of the SN were depleted of their glycogen stores by repetitive stimulation of the nerve. Transverse sections were taken from treated and control muscles and stained with the periodic-acid Schiffs stain. A n example from such an experiment is shown in Fig. 3. Table 3 summarises the results and shows that in normal soleus muscles and in those treated with a single BTX implant stimulation of the L4 SN depleted approximately 30% of the muscle fibres (expressed as a percentage of the contralateral control
muscle). In 2 animals where the soleus muscle was treated with 2 BTX implants, the majority of the muscle fibres were innervated by axons in the L4 SN, and only very few were supplied by the L5 SN. This may reflect the reduced number of motor units in the L5 SN after BTX treatment as shown in Table 2. The massive reduction in muscle fibres supplied by the L5 SN in the 2 animals treated with 2 BTX implants is probably due to a loss of motoneurones that exit through the L5 SN. However, in these 2 animals the number of muscle fibres supplied by the L4 SN is higher than usual, indicating that the territory occupied by axons in the L4 SN expanded. The results taken together show that transient muscle paralysis of the soleus muscle shortly after birth almost exclusively affects motor units that have their axons in the L5 SN.
4. Discussion
Fig. 3. This figure shows an example of a transverse section (10 /zm) of an adult soleus muscle, in which the muscle fibres innervatedby either the L4 or L5 spinal nerve were depleted of their glycogen stores by repetitive stimulation of the nerve. Muscle sections were then subsequently stained with the perio~Lc-acid Schiff stain. The number of muscle fibres innervated by each of the spinal nerves could then be distinguishedand countedusing a camera lucida device attached to a light microscope. Scale bar = 100 /xm.
The results of the present study confirm earlier findings showing that treatment of the rat soleus muscle with BTX shortly after birth leads to muscle weakness, and a loss of motor units and muscle fibres [4,7]. After repeated application of BTX, as many as 50% of the motor units are lost, and the force output of the treated muscle is correspondingly reduced [4]. Treatment of neonatal muscles with BTX in the manner described in this study prevents neuromuscular transmission for at least 8 - 1 0 days [7]. This transient interruption of nerve-muscle interaction during this early period of postnatal development is sufficient to affect a population of motoneurones causing them to die. Whether the motoneurones that are affected by this disruption of neuromuscular interaction differ in any way from those that survive is not clear. Since the maturation of the spinal cord proceeds in a cranio-caudal direction, and bearing in mind that this maturation is not complete at the time of birth, it is possible that the motoneurones in the more caudal segments are less mature. It is likely that these motoneurones are therefore more susceptible to target deprivation, since previous work has shown that imma-
286
L. Greensmith et al. / Developmental Brain Research 98 (1997) 281-286
ture m o t o n e u r o n e s are more vulnerable to nerve injury than motoneurone of more mature animals [3,5,10,11,18,19]. The results of the present study support this possibility, and show that while the m o t o n e u r o n e s in the more cranial segments are apparently unaffected b y temporary disruption of n e u r o m u s c u l a r interaction shortly after birth, m o t o n e u r o n e s in the more caudal s e g m e n t are severely and selectively affected, so that m a n y of them die as a result of the transient neonatal m u s c l e paralysis. It could be argued that the position of the B T X i m p l a n t on the soleus muscle m a y preferentially affect specific groups of muscle fibres which are closer to the silicon rubber strip, and that these muscle fibres might be supplied by m o t o n e u r o n e s from the more caudal segments of the spinal cord. However, several findings argue against this possibility: (i) our experiments where glycogen depletion of the n o r m a l adult soleus muscle was carried out by stimulating either the L4 or the L5 spinal nerves, show that the muscle fibres innervated by either SN are r a n d o m l y intermingled, and (ii) at the time w h e n B T X treatment is carried out in the neonatal rats, i.e. shortly after birth, almost all muscle fibres are innervated b y axons from both spinal nerves due to the p o l y n e u r o n a l innervation that developing muscle fibres receive [1,15,17]. It is therefore unlikely that the m u s c l e paralysis would selectively affect muscle fibres supplied b y the L5 spinal nerve, causing a selective loss of m o t o n e u r o n e s whose axons leave the spinal cord in the L5 spinal nerve. Nevertheless, we c a n n o t exclude the possibility that the n e u r o m u s c u l a r j u n c t i o n s formed by nerve terminals of axons from the L5 SN are affected differently by treatment with BTX, and that those from axons in the L4 SN are in some way more resistant to B T X treatment and the consequent disruption of nerve:muscle interaction. At this stage, however, it is impossible to say whether this is indeed the case. Therefore, the most likely explanation of the results of this study is that due to the cranio-caudal sequence of maturation of the spinal cord, the more caudally situated motoneurones are less mature than those located more cranially, and as a consequence are more vulnerable to deprivation of n e u r o m u s c u l a r interaction during early postnatal development.
Acknowledgements W e would like to thank The W e l l c o m e Trust, The M u s c u l a r Dystrophy Group of Great Britain and The EEC for their support. W e are also grateful to Mr. J. D i c k and Mrs. D. Bartram for their excellent technical and secretarial assistance.
References [1] Brown, M.C., Jansen, J.K. and Van Essen, D., Polyneuronal innervation of skeletal muscle in newborn rats and its elimination during maturation, J. Physiol. Lond.. 261 (1976) 387-422. [2] Greensmith, L., Dick, J., EmanueL A.O. and Vrbova, G., Induction of transmitter release at the neuromuscular junction prevents motoneuron death after axotomy in neonatal rats, Neuroscience, 71 (1996) 213-220. [3] Greensmith, L., Hasan, H.I. and Vrbova, G., Nerve injury increases the susceptibility of motoneurons to N-methyl-D-aspartate-induced neurotoxicity in the developing rat, Neuroscience, 58 (1994) 727733. [4] Greensmith, L., Hind, A.H. and Vrbova, G., Transient disruption of nerve-muscle interaction shortly after birth permanently alters the development of the rat soleus muscle, Dev. Brain Res., 94 (1996) 152-158. [5] Greensmith, L., Mentis, G.Z. and Vrbova, G., Blockade of Nmethyl-D-aspartate receptors by MK-801 (dizocilpine maleate) rescues motoneurones in developing rats, Dev. Brain Res., 81 (1994)
162-170. [6] Greensmith, L. and Vrbova, G., Neuromuscular contacts in the developing rat soleus depend on muscle activity, Dev. Brain Res., 62 (1991) 121-129. [7] Greensmith, L. and Vrbova, G., Alterations of nerve-muscle interaction during postnatal development influence motoneurone survival in rats, Dev. Brain Res., 69 (1992) 125-131. [8] Greensmith, L. and Vrbova, G., Motoneurone survival: A functional approach, Trends Neurosci., 19 (1996) 450-455. [9] Kugelberg, E., The motor unit: anatomy and histochemical functional correlations, Riv. Patol. Nerv. Ment., 97 (1976) 251-258. [10] Lowrie, M.B., Krishnan, S. and Vrbova, G., Recovery of slow and fast muscles following nerve injury during early post-natal development in the rat, J. Physiol. Lond., 331 (1982) 51-66. [11] Lowrie, M.B., Krishnan, S. and Vrbova, G., Permanent changes in muscle and motoneurones induced by nerve injury during a critical period of development of the rat, Brain Res., 428 (1987) 91-101. [12] Lowrie, M.B., O'Brien, R.A.D. and Vrbova, G., The effect of altered peripheral field on motoneurone function in developing rat soleus muscles, J. Physiol., 368 (1985) 513-524. [13] Lowrie, M.B. and Vrbova, G., Dependence of postnatal motoneurones on their targets: review and hypothesis, Trends Neurosci., 15 (1992) 80-84. [14] Nachlas, M.M., Tson, K.C., DeSouza, E., Chang, C.H. and Seligman, A.M., Cytochemical demonstration of succinate dehydrogenase by the use of a new p-nitrophenyl substituted dietrazole, J. Histochem. Cytochem., 5 (1957) 420-436. [15] O'Brien, R.A., Ostberg, A.J. and Vrbova, G., Observations on the elimination of polyneuronal innervation in developing mammalian skeletal muscle, J. Physiol. Lond., 282 (1978) 571-582. [16] Pearse, A.G.E., Histochemistry, 2nd Edition, Little Brown, Boston, 1960. [17] Redfern, P.A., Neuromuscular transmission in new-born rats, J. Physiol. Lond., 209 (1970) 701-709. [18] Romanes, G.J., Motor localisation and the effects of nerve injury on the ventral horn cells of the spinal cord, J. Anat., 80 (1946) 11-131. [19] Schmalbruch, H., Motoneurone death after sciatic nerve section in rats, J. Comp. Neurol., 224, 252-258.
ELSEVIER
DevelopmentalBrain Research 98 (1997) 281-286
Research report
Neonatal paralysis of the rat soleus muscle selectively affects motoneurones from more caudal segments of the spinal cord Linda Greensmith, Angela Hind, Gerta Vrbovfi * Department of Anatomy and Developmental Biology, Centre for Neuroscience, University College London, Gower Street, London, WCIE 6BT, UK Accepted 29 October 1996
Abstract Transient paralysis of the rat soleus muscle shortly after birth leads to permanent muscle weakness, loss of muscle fibres and the death of motoneurones. The soleus muscle receives its innervation from motoneurones whose axons exit the spinal cord either via the L4 ventral ramus or in the more caudal part of the cord via the L5 ventral ramus. Whether both populations of motoneurones are equally affected by neonatal paralysis was studied here. In soleus muscles paralysed with a-bungarotoxin shortly after birth and examined 8-10 weeks later, there is no loss of force or muscle fibres in the part of the muscle supplied by axons in the L4 ventral ramus. Loss of force and muscle fibre numbers occurs only in the part of the muscle supplied by axons in the L5 ventral ramus. In a normal adult soleus 30.3 + 2.4% of muscle force is produced by stimulating the L4 ventral ramus and 69.0 ___5.5% by stimulating the L5 ventral ramus. In soleus muscles treated with a-bungarotoxin 28 _ 1.4% of the force produced by the contralateral control soleus was generated by axons in the L4 ventral ramus and only 20.3 + 5.6% by stimulating the L5 spinal nerve. The number of muscle fibres supplied by either ventral ramus in control and experimental muscles confirmed that the decrease of force after treatment with a-bungarotoxin can be accounted for by loss of muscle fibres supplied by axons in the L5 ventral ramus. The reduced force production and muscle fibre numbers was due to a selective loss of motoneurones that had their axons in the L5 ventral ramus. The number of axons to soleus in the L4 ventral ramus was 9.3 + 0.7 in controls and 10.3 _ 0.9 in the experimental animals, whereas the L5 ventral ramus contained 17.2 + 0.7 in controls and only 4.7 + 1.7 in the experimental animals. Thus paralysis of the soleus muscle at birth selectively affects motoneurones in the more caudal part of the spinal cord, suggesting that the more cranial motoneurones are more mature and less likely to be influenced by lack of neuromuscular interaction at the time of birth. Keywords: Motoneurone;Nerve:muscleinteraction;Development;Paralysis
1. Introduction It is well established that during the early postnatal period motoneurones remain dependant upon contact with their target muscle for l~heir survival. Following nerve injury at birth, more than 70% of motoneurones to the soleus muscle die [5,11,18]. This death of target-deprived motoneurones is a consequence of the lack of functional interaction between the motoneurone and its target muscle, since disruption of nerve:muscle interaction during this critical period of development also results in motoneurone death. We have previously reported that a brief, transient paralysis of the soleus muscle in neonatal rats results in the death of a large proportion of motoneurones to the soleus
* Corresponding author. Fax: +44 (171) 380 7349. E-mail: [email protected]
muscle [7]. Moreover, the soleus muscles paralysed during this critical period of postnatal development remained permanently weak, due to the loss of a proportion of their muscle fibres [4]. The extent of the reduction of force and loss of muscle fibres correlated with the decrease in motoneurone numbers. When the muscle was paralysed for only a short period of time, around 30% of motoneurones to the soleus muscle died and there was an equivalent reduction in muscle force. Prolonging the duration of the paralysis resulted in a greater loss of both motoneurones and muscle fibres. Thus it appears that following disruption of neuromuscular interaction shortly after birth, the main reason for the consequent muscle weakness was the loss of motoneurones. However, in these experiments, following neonatal muscle paralysis, not all the motoneurones to the soleus muscle died, so that some of the motoneurones possessed properties that rendered them resistant to the effects of target deprivation. Therefore there
0165-3806/97/$17.00 Copyright © 1997 Elsevier Science B.V. All rights reserved. PII SO 165 -3 806(96)00200-3
282
L. Greensmithet al./ DevelopmentalBrain Research 98 (1997) 281-286
are clearly differences in the responses of motoneurones to disruption of neuromuscular interaction. Neuromuscular interaction in neonatal rats can be disrupted by blocking the postsynaptic acetylcholine receptor (ACh) with the snake toxin, c~-Bungarotoxin (BTX). This treatment of the neuromuscular junction of the soleus muscle in newborn rats, results in muscle paralysis that lasts for 2 - 3 days, and this paralysis can be prolonged by additional treatments with BTX [7]. In this study we wished to establish which motoneurones are more likely to be affected by transient neonatal paralysis. The soleus muscle receives its innervation via two ventral roots: 30% of motor axons to soleus exit through the L4 spinal nerve (SN) and these are axons from motoneurones situated in the more cranial segment of the cord; 70% of motor axons are in the L5 SN and most of these are from the more caudal part of the soleus motor pool. In view of the cranio-caudal progression of the maturation of the spinal cord, motoneurones in the L4 segment might be expected to be more mature than those in the L5 part of the soleus motoneurone pool. It is well known that the susceptibility of motoneurones to separation from their target declines rapidly with age [18] (for reviews see [8,13]). For example, we have previously found that sciatic nerve crush in newborn rats results in the death of as many as 90% of motoneurones [5], injury at 3 days causes the loss of up to 80% [2], but the same insult carded out in 5-day-old rats causes little if any motoneurone death [3,10]. It is therefore possible that the more cranial, mature motoneurones in the L4 segment are less likely to be affected by temporary loss of interaction with the muscle fibres they supply. The present study was designed to examine this possibility.
silicon rubber (Dow Coming, 3140 RTV), weighing 0.3 mg and containing approximately 7 / z g BTX and 120 /zg NaC1 were implanted into the hindlimb of newborn rats. Strips weighing 0.75 mg and containing 16 /xg BTX and 240 /~g NaC1 were implanted into the hindlimbs of 3 day old rats. Control implants of similar weights were implanted into control animals and contained 120 /zg and 240 /xg NaC1 at each of the respective ages. 2.3. Isometric tension recordings When the rats were 2 - 1 0 months old, they were anaesthetised with chloral hydrate (4%, 1 m l / 1 0 0 g body weight, i.p.) and the soleus muscles in both the treated and contralateral control legs were prepared for in vivo assessment of muscle force. The soleus muscle in each leg was dissected free of its surrounding tissue. Both legs were rigidly secured to a table and the distal tendons attached to isometric force transducers (Dynameter UFI, Devices) via silk threads. The sciatic nerve was exposed in the leg, isolated from the surrounding tissue and followed proximally to the point where the L4 and L5 spinal nerves (SN) join to form the sciatic. The spinal nerves were then dissected free as far proximally as possible, sectioned and ligated. In the contralateral control leg, the soleus nerve was dissected. All other muscles of the hindlimbs were then denervated and their distal tendons severed. The length of the soleus muscles in both legs was adjusted to that at which the muscles developed maximal twitch tension. Isometric contractions were elicited by stimulating the cut end of either the L4 or L5 SN, or the motor nerve to the soleus muscle, using a pulse width of 0.02 ms. To elicit tetanic contraction the muscles were stimulated at 20, 40 and 80 Hz for 800 ms.
2. Materials and methods
2.4. Motor unit numbers
2.1. Surgical procedures
To estimate the total number of motor units in the treated and contralateral control soleus muscles the motor nerve to soleus was stimulated by single pulses every 4 s. To estimate the number of motor axons in each SN to the treated soleus muscle, the distal end of each SN was individually stimulated and the twitch tension of soleus recorded. The stimulus intensity applied to the nerve was gradually increased to obtain stepwise increments of twitch tension as individual motor axons with increasing stimulus thresholds are recruited. The number of stepwise increments was counted.
Under halothane anaesthesia and sterile conditions, a small silicon rubber strip, containing either o~-bungarotoxin (BTX, Sigma) or NaC1, was implanted alongside the soleus muscle in the right hindlimb of neonatal Wistar rats, within 6 - 1 2 h of birth. The contralateral leg was left unoperated and served as a control. When the pups had fully recovered from the anaesthesia they were returned to their mothers. Some of these operated animals received a second silicon strip 3 days after the initial operation. Under the same conditions as above, the first implant was identified and removed and replaced with a larger implant containing either BTX or NaC1, as appropriate. On recovery from the anaesthesia these animals were returned to their mothers and reared as normal. 2.2. Preparation of implants The preparation of BTX containing implants has previously been described [6]. Silicon strips prepared from inert
2.5. Glycogen depletion In some of the experiments in which the maximum tetanic tension of fibres innervated by a specific SN had been estimated, the technique of glycogen depletion was employed as a means of identifying the number and location of the muscle fibres innervated by axons emanating from either of the SN. Glycogen within the muscle fibres
L. Greensmith et al. / Developmental Brain Research 98 (1997) 281-286
283
was depleted by repetitive stimulation of the muscle at 40 Hz for 250 ms, via either the L4 or the L5 SN. The muscles were usually stimulated for 4 5 - 6 0 min, or until the tetanic contraction produced in response to stimulation had decreased to approximately 5% of the original. 2.6. Muscle histology
After the tension experiments had been completed, the control and treated soleus muscles were dissected out from the rats under terminal chloral hydrate anaesthesia. The muscles were then stretched to a length similar to that in the leg and the control and treated muscle from each animal was mounted side by side and frozen in melting isopentane. To determine the number of muscle fibres, transverse cryostat sections (10 /xm) were collected onto gelatinised slides and stained for succinate dehydrogenase (SDH) [14]. Sections taken from muscles that had undergone glycogen depletion of fibres innervated by axons of either the L4 or the L5 SN were processed for the periodic acid-Schiff (PAS) reaction for glycogen [16]. This method demonstrates the presence of glycogen in muscle fibres. The number of muscle fibres innervated by each SN could then be distinguished, depending on their staining with the PAS stain. Those fibres which had been depleted of their glycogen stores by repetitive stimulation via their innervating SN did not react with the PAS stain. The number of muscle fibres in each soleus muscle was counted using a camera lucida attached to a light microscope. In those muscle in which the muscle fibres innervated by one of the SN lind been depleted of glycogen, both the total number of muscle fibres and the number innervated by each of the SN was counted.
3. Results 3.1. Tetanic tension in response to stimulation of either the L4 or L5 spinal nerve
The contribution of the: tension elicited by stimulating either the L4 or the L5 SN to the total force output (maximum tetanic tension or MTT) of the normal soleus muscle of adult rats was assessed in 7 animals aged between 2 and 10 months. An example of the tension recording from such an experiment is shown in Fig. la. Table 1 summarises the results and shows that in normal soleus muscles stimulation of the IA SN elicited 30.7% ( _ 2.7 S.E.M., n = 7) of the total muscle force, and stimulation of the L5 SN elicited 69.5% (+__7.5 S.E.M., n = 7). These results are consistent with values reported by other authors [12]. Soleus muscles that had been transiently paralysed by treatment with a single BTX implant shortly after birth were studied next. Fig. lb shows an example from an
b
25g[ 100m~ Fig. 1. Examples of isometric muscle contractions from adult soleus muscles are shown. Trace (a) is from a normal, control soleus muscle, (b) from a muscle treated with a single BTX implant at birth, and (c) is from a muscle treated with 2 BTX implants, at birth and 3 days of age. The maximum tetanic tension elicited by stimulating either the entire sciatic nerve, or the IA or L5 spinal nerve is illustrated.
experiment in which the MTT of adult soleus muscles was recorded in response to stimulation of either the L4 or L5 SN or the entire sciatic nerve. The results from the experiments in this group are summarised in Table 1, which shows that the soleus muscles treated with BTX at birth produced 40.5% ( _ 2.6 S.E.M., n = 7) of their MTT in response to stimulation of the L4 SN and 63% ( + 2 . 9 S.E.M., n = 7) in response to stimulation of the L5 SN. Examples of records taken from a soleus muscle that had been treated shortly after birth with 2 successive implants containing BTX and which were therefore paralysed for a longer period of time, are shown in Fig. lc. Table 1 summarises the results and shows that adult soleus muscles treated in this way during the early postnatal period develop 59.4% ( + 7.4 S.E.M., n = 4) of their total force in response to stimulation of their L4 SN, and 39.3% ( _ 7.4 S.E.M., n = 4) in response to stimulation of the L5 SN. Thus after treatment with BTX shortly after birth, the proportional contribution of the L4 and L5 SN to the force output of the adult soleus muscle has changed. Fig. 2 illustrates this result. Since the BTX-treated soleus muscles are permanently weaker than untreated control muscles and have fewer
L. Greensmith et al. / Developmental Brain Research 98 (1997) 281-286
284
A
90 80.
BL4 --i--
raL5
70,
Motor unit number
60.
1.,4
L5
9.3 5:0.7 10.5 + 1.0 10.3 + 0.9
17.2 + 0.7 15.0+0.9 4.7 ___1.7
50
.~ 40 "" 30, [- 20.
Control (7) 1 BTX (7) 2 BTX (4)
J::
Control (7)
NaCI (4)
1 BTX
2 BTX
(7)
(4)
The mean number of motor units ( 5: S.E.M.) from adult soleus muscles in response to stimulation of either the L4 or L5 spinal nerve is shown, in control animals and animals which received either a single BTX implant at birth, or 2 BTX implants, at birth and 3 days, and were examined 8-10 weeks later. The number of animals in each group is shown in parentheses.
B
9
Table 2 Number of motor units in adult soleus after neonatal treatment with BTX
120
IlL4
100
QL5
N
80
60
Fig. 2 summarise the results and show that the reduced force seen in the muscles treated with BTX can be entirely accounted for by the reduced contribution of the L5 SN to the total force output of the muscle.
40
3.2. Motor unit numbers
20 0 Control (7)
NaCI (4)
1 BTX
2 BTX
(7)
(4)
Fig. 2. The block diagrams illustrate the maximum tetanic tension elicited in adult animals from normal, control soleus muscles and those treated with either NaC1, a single BTX implant at birth, or 2 BTX implants at birth and 3 days of age, in response to stimulation of either the L4 or L5 spinal nerve, and expressed (A) as a % of the total muscle force, and (B) as a % of the total force of the contralateral control soleus muscle. The number of animals in each group is shown in parentheses, and the error bars represent the S.E.M..
muscle fibres (see [4]), it is possible that this alteration in the contribution to force output produced by stimulation of either SN is caused by a disproportionately greater loss of force from the part of the muscle supplied by the L5 spinal nerve. Therefore, in each animal the MTT obtained by stimulating either spinal nerve was expressed as a percentage of the total force produced by stimulating the soleus motor nerve in the contralateral control leg. Table 1 and
We have previously shown that following transient paralysis of the soleus muscle at birth, there is a loss of motoneurones [7] and a reduction in the number of surviving motor axons in the soleus nerve [4]. Whether this loss of motoneurones is evenly distributed between the L4 and L5 lumbar segments was studied next. Table 2 summarises the results of the number of motor units obtained by recording the stepwise increments of single twitch tension from the soleus muscle in response to stimulation of either the L4 or L5 SN by stimuli of increasing intensity. The table shows that the mean number of axons in the L4 SN is approximately the same in all the experimental groups, but the number of axons in the L5 SN is lower in animals treated with BTX implants at birth, and this decrease is much greater in the group of animals treated with 2 BTX implants. These results indicate that paralysis of the soleus muscle at birth affects mainly those motoneurones that send their axons through the L5 SN and that those that leave the spinal cord through the more cranial nerve are spared.
Table 1 Maximum tetanic tension of soleus muscle after neonatal treatment with BTX Treatment
Maximum tetanic tension Total muscle force (%op/con)
Control (7) NaC1 (4) 1 BTX (7) 2 BTX (4)
98.0 + 6.7 117 + 7.3 73 + 4.5 49.5 -I- 4.3
Stimulation of L4 or L5 SN (% total force)
Stimulation of L4 or L5 SN (% total force of control soleus)
L4
L4
30.7 20.0 40.5 59.4
L5 + + _ +
2.7 10.2 2.6 7.4
69.0 80.0 63.0 39.3
_ 3.5 _ 8.6 + 2.9 -I- 7.4
30.3 27.4 28.4 28.7
L5 + + + +
2.4 11 3.5 1.4
69.0 95.6 46.1 20.3
-I- 5.5 + 8.1 + 3.9 5:5.6
The mean values ( ___S.E.M.) of muscle force (MTT) developed by control soleus muscles of adult animals, and by soleus muscles that were treated with NaC1, a BTX implant at birth or 2 BTX implants, at birth and 3 days later, are expressed as a percentage of the control muscle. In addition, the MTT elicited by stimulation of either the L4 or L5 spinal nerve is shown. These values are expressed as a percentage of the MTT developed by the experimental muscle, and as a percentage of the MTT of the contralateral control soleus. The number of animals in each group is shown in parentheses.
L. Greensmith et aL/ Developmental Brain Research 98 (1997) 281-286
285
Table 3 Number of soleus muscle fibres innervatedby the IA and L5 spinal nerves Treatment
Control (7) NaC1 (4) 1 BTX (7) 2 BTX (2)
Number of musclefibres
% Of total numberof musclefibres
% Of total numberof muscle fibres in control soleus
L4
L5
L4
L5
L4
L5
812 + 125 786 + 149 753 ± 102 1705
1807 ___102 2007 + 76 1555 + 317 112
30.8 + 3.7 28.0 + 4.4 30.0 ± 7.1 94
69 + 3.2 72 + 4.4 64 _+7.1 6
20.8 + 3.6 27.3 + 3.5 30 _ 5.5 64
69.3 + 2.0 69.8 + 2.0 56.2 + 7.6 4
Glycogen was depleted from muscle fibres suppliedby axons in the LA spinal nerve by repeated stimulation.This permitted identificationof the muscle fibres innervatedby axons in either the L4 or L5 spinal nerve. The mean number of glycogendepleted muscle fibres ( + S.E.M.) innervatedby the L4 and L5 spinal nerves is shown for control soleus muscles and those treated with either NaC1, 1BTX implant at birth, or 2 BTX implants at birth and 3 days of age. The results are also expressed as a percentage of the total number of muscle fibres present, and as a percentage of the total numberof muscle fibres in the contralateralcontrol soleus muscles. The number of animals in each group is given in parentheses.
3.3. M u s c l e f i b r e n u m b e r s
The next set of experiments was carded out in order to establish the numbers of muscle fibres supplied by axons in the L4 or the L5 SN respectively in both normal soleus muscles and those treated with a single or double BTX implant. The number of muscle fibres innervated by each of the SN was investigated using the technique of glycogen depletion [9]. The muscle fibres innervated by one of the SN were depleted of their glycogen stores by repetitive stimulation of the nerve. Transverse sections were taken from treated and control muscles and stained with the periodic-acid Schiffs stain. A n example from such an experiment is shown in Fig. 3. Table 3 summarises the results and shows that in normal soleus muscles and in those treated with a single BTX implant stimulation of the L4 SN depleted approximately 30% of the muscle fibres (expressed as a percentage of the contralateral control
muscle). In 2 animals where the soleus muscle was treated with 2 BTX implants, the majority of the muscle fibres were innervated by axons in the L4 SN, and only very few were supplied by the L5 SN. This may reflect the reduced number of motor units in the L5 SN after BTX treatment as shown in Table 2. The massive reduction in muscle fibres supplied by the L5 SN in the 2 animals treated with 2 BTX implants is probably due to a loss of motoneurones that exit through the L5 SN. However, in these 2 animals the number of muscle fibres supplied by the L4 SN is higher than usual, indicating that the territory occupied by axons in the L4 SN expanded. The results taken together show that transient muscle paralysis of the soleus muscle shortly after birth almost exclusively affects motor units that have their axons in the L5 SN.
4. Discussion
Fig. 3. This figure shows an example of a transverse section (10 /zm) of an adult soleus muscle, in which the muscle fibres innervatedby either the L4 or L5 spinal nerve were depleted of their glycogen stores by repetitive stimulation of the nerve. Muscle sections were then subsequently stained with the perio~Lc-acid Schiff stain. The number of muscle fibres innervated by each of the spinal nerves could then be distinguishedand countedusing a camera lucida device attached to a light microscope. Scale bar = 100 /xm.
The results of the present study confirm earlier findings showing that treatment of the rat soleus muscle with BTX shortly after birth leads to muscle weakness, and a loss of motor units and muscle fibres [4,7]. After repeated application of BTX, as many as 50% of the motor units are lost, and the force output of the treated muscle is correspondingly reduced [4]. Treatment of neonatal muscles with BTX in the manner described in this study prevents neuromuscular transmission for at least 8 - 1 0 days [7]. This transient interruption of nerve-muscle interaction during this early period of postnatal development is sufficient to affect a population of motoneurones causing them to die. Whether the motoneurones that are affected by this disruption of neuromuscular interaction differ in any way from those that survive is not clear. Since the maturation of the spinal cord proceeds in a cranio-caudal direction, and bearing in mind that this maturation is not complete at the time of birth, it is possible that the motoneurones in the more caudal segments are less mature. It is likely that these motoneurones are therefore more susceptible to target deprivation, since previous work has shown that imma-
286
L. Greensmith et al. / Developmental Brain Research 98 (1997) 281-286
ture m o t o n e u r o n e s are more vulnerable to nerve injury than motoneurone of more mature animals [3,5,10,11,18,19]. The results of the present study support this possibility, and show that while the m o t o n e u r o n e s in the more cranial segments are apparently unaffected b y temporary disruption of n e u r o m u s c u l a r interaction shortly after birth, m o t o n e u r o n e s in the more caudal s e g m e n t are severely and selectively affected, so that m a n y of them die as a result of the transient neonatal m u s c l e paralysis. It could be argued that the position of the B T X i m p l a n t on the soleus muscle m a y preferentially affect specific groups of muscle fibres which are closer to the silicon rubber strip, and that these muscle fibres might be supplied by m o t o n e u r o n e s from the more caudal segments of the spinal cord. However, several findings argue against this possibility: (i) our experiments where glycogen depletion of the n o r m a l adult soleus muscle was carried out by stimulating either the L4 or the L5 spinal nerves, show that the muscle fibres innervated by either SN are r a n d o m l y intermingled, and (ii) at the time w h e n B T X treatment is carried out in the neonatal rats, i.e. shortly after birth, almost all muscle fibres are innervated b y axons from both spinal nerves due to the p o l y n e u r o n a l innervation that developing muscle fibres receive [1,15,17]. It is therefore unlikely that the m u s c l e paralysis would selectively affect muscle fibres supplied b y the L5 spinal nerve, causing a selective loss of m o t o n e u r o n e s whose axons leave the spinal cord in the L5 spinal nerve. Nevertheless, we c a n n o t exclude the possibility that the n e u r o m u s c u l a r j u n c t i o n s formed by nerve terminals of axons from the L5 SN are affected differently by treatment with BTX, and that those from axons in the L4 SN are in some way more resistant to B T X treatment and the consequent disruption of nerve:muscle interaction. At this stage, however, it is impossible to say whether this is indeed the case. Therefore, the most likely explanation of the results of this study is that due to the cranio-caudal sequence of maturation of the spinal cord, the more caudally situated motoneurones are less mature than those located more cranially, and as a consequence are more vulnerable to deprivation of n e u r o m u s c u l a r interaction during early postnatal development.
Acknowledgements W e would like to thank The W e l l c o m e Trust, The M u s c u l a r Dystrophy Group of Great Britain and The EEC for their support. W e are also grateful to Mr. J. D i c k and Mrs. D. Bartram for their excellent technical and secretarial assistance.
References [1] Brown, M.C., Jansen, J.K. and Van Essen, D., Polyneuronal innervation of skeletal muscle in newborn rats and its elimination during maturation, J. Physiol. Lond.. 261 (1976) 387-422. [2] Greensmith, L., Dick, J., EmanueL A.O. and Vrbova, G., Induction of transmitter release at the neuromuscular junction prevents motoneuron death after axotomy in neonatal rats, Neuroscience, 71 (1996) 213-220. [3] Greensmith, L., Hasan, H.I. and Vrbova, G., Nerve injury increases the susceptibility of motoneurons to N-methyl-D-aspartate-induced neurotoxicity in the developing rat, Neuroscience, 58 (1994) 727733. [4] Greensmith, L., Hind, A.H. and Vrbova, G., Transient disruption of nerve-muscle interaction shortly after birth permanently alters the development of the rat soleus muscle, Dev. Brain Res., 94 (1996) 152-158. [5] Greensmith, L., Mentis, G.Z. and Vrbova, G., Blockade of Nmethyl-D-aspartate receptors by MK-801 (dizocilpine maleate) rescues motoneurones in developing rats, Dev. Brain Res., 81 (1994)
162-170. [6] Greensmith, L. and Vrbova, G., Neuromuscular contacts in the developing rat soleus depend on muscle activity, Dev. Brain Res., 62 (1991) 121-129. [7] Greensmith, L. and Vrbova, G., Alterations of nerve-muscle interaction during postnatal development influence motoneurone survival in rats, Dev. Brain Res., 69 (1992) 125-131. [8] Greensmith, L. and Vrbova, G., Motoneurone survival: A functional approach, Trends Neurosci., 19 (1996) 450-455. [9] Kugelberg, E., The motor unit: anatomy and histochemical functional correlations, Riv. Patol. Nerv. Ment., 97 (1976) 251-258. [10] Lowrie, M.B., Krishnan, S. and Vrbova, G., Recovery of slow and fast muscles following nerve injury during early post-natal development in the rat, J. Physiol. Lond., 331 (1982) 51-66. [11] Lowrie, M.B., Krishnan, S. and Vrbova, G., Permanent changes in muscle and motoneurones induced by nerve injury during a critical period of development of the rat, Brain Res., 428 (1987) 91-101. [12] Lowrie, M.B., O'Brien, R.A.D. and Vrbova, G., The effect of altered peripheral field on motoneurone function in developing rat soleus muscles, J. Physiol., 368 (1985) 513-524. [13] Lowrie, M.B. and Vrbova, G., Dependence of postnatal motoneurones on their targets: review and hypothesis, Trends Neurosci., 15 (1992) 80-84. [14] Nachlas, M.M., Tson, K.C., DeSouza, E., Chang, C.H. and Seligman, A.M., Cytochemical demonstration of succinate dehydrogenase by the use of a new p-nitrophenyl substituted dietrazole, J. Histochem. Cytochem., 5 (1957) 420-436. [15] O'Brien, R.A., Ostberg, A.J. and Vrbova, G., Observations on the elimination of polyneuronal innervation in developing mammalian skeletal muscle, J. Physiol. Lond., 282 (1978) 571-582. [16] Pearse, A.G.E., Histochemistry, 2nd Edition, Little Brown, Boston, 1960. [17] Redfern, P.A., Neuromuscular transmission in new-born rats, J. Physiol. Lond., 209 (1970) 701-709. [18] Romanes, G.J., Motor localisation and the effects of nerve injury on the ventral horn cells of the spinal cord, J. Anat., 80 (1946) 11-131. [19] Schmalbruch, H., Motoneurone death after sciatic nerve section in rats, J. Comp. Neurol., 224, 252-258.