- Email: [email protected]
Motoneurons innervating partially denervated rat hindlimb muscles remain susceptible to axotomy-induced cell death
Pergamon
PII:
Neuroscience Vol. 86, No. 1, pp. 291–299, 1998 Copyright ? 1998 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/98 $19.00+0.00 S0306-4522(98)00037-2
MOTONEURONS INNERVATING PARTIALLY DENERVATED RAT HINDLIMB MUSCLES REMAIN SUSCEPTIBLE TO AXOTOMY-INDUCED CELL DEATH D. I. HARDING, L. GREENSMITH, P. N. ANDERSON and G. VRBOVA u* Department of Anatomy and Developmental Biology, The Centre for Neuroscience, University College London, Gower Street, London WC1E 6BT, U.K. Abstract––Tibialis anterior and extensor digitorum longus muscles were partially denervated by cutting the L4 spinal nerve in three-day-old rats. The ultrastructure of the intact axons to these muscles in the L5 spinal nerve was examined in nine-day-old rats. In the control L5 spinal nerve, myelinated and unmyelinated axons were intermingled throughout the cross-section of the nerve, while on the operated side the nerve contained areas with predominantly small unmyelinated immature axons. The number of motoneurons innervating the partially denervated muscles was established by retrograde labelling with Diamidino Yellow. In nine- and 21-day-old rats, the number of labelled motoneurons on the partially denervated side, expressed as a percentage of the control side, was 26.1&5.5% and 20.7&3.0%, respectively. The response of these uninjured motoneurons to axotomy was tested. The axons of the motoneurons to the partially denervated muscles were crushed at nine days and the numbers of labelled motoneurons in the spinal cord of these rats counted at 21 days of age. Only 4.9&2.0% labelled motoneurons were seen on the operated side, as opposed to 20.7&3.0% present in animals without sciatic nerve injury. In normal animals, nerve injury at nine days does not cause motoneuron death. Thus, motoneurons to partially denervated muscles (i) have axons with several immature features and (ii) remain susceptible to axotomy-induced death for much longer than normal. ? 1998 IBRO. Published by Elsevier Science Ltd. Key words: axon ultrastructure, motoneuron death, neuronal maturation, target interaction.
Motoneurons of newborn animals die following injury to their axons. In rats, sciatic nerve injury at birth leads to the death of 80–90% of motoneurons, whilst the same insult carried out at five days of age results in little, if any, motoneuron death.20,22 Thus, during the first five days of life motoneurons undergo fundamental changes that enable them to withstand nerve injury. There is some evidence to suggest that this transition, from a motoneuron that is dependent upon contact with its target to one that can survive separation, is induced by functional interaction with the muscle during this critical developmental period. For example, motoneuron death was observed in the spinal cord of rats where active interaction of the motor nerve terminals with the muscle fibres was prevented by paralysing the muscle with á-bungarotoxin during the first few days of postnatal development.13 In these experiments, no injury to the motor axons was inflicted, but nevertheless a substantial loss of motoneurons occurred. These results indicate that some event that is related to the interaction between the nerve terminals and their muscle fibres is involved in the changes that occur in the motoneuron that later enables it to survive target deprivation. *To whom correspondence should be addressed. Abbreviations: DiY, Diamidino Yellow; EDL, extensor digitorum longus; SN, spinal nerve; TA, tibialis anterior. 291
Several proposals have been made as to the precise role of the target in the survival and development of motoneurons, but the exact nature of the target on the developing motoneuron is still not clear. Thus, a better understanding of the nature of the influence of the target on motoneuron survival has yet to be established. The early postnatal period marks a time when motoneurons are not only dependent upon their target for survival, but also a time when both the developing nerve terminals and the postsynaptic membrane are undergoing changes as they mature. One of these changes is the arrest of growth of the nerve terminal and an increase in transmitter release triggered by its encounter with muscle fibres.12,30 This change in the nerve terminal from a growing into a transmitting structure is reflected within the cell body, so that the motoneuron also changes from a growing into a transmitting cell. It has been proposed that the motoneuron must undergo this transition within a specific period of development if it is to mature appropriately and survive within the developing spinal cord.16 According to this hypothesis, motoneurons that are maintained in a growing mode for a prolonged time during their development fail to mature and are unable to survive the effects of increased excitatory inputs within the developing spinal cord.23 Several
292
D. I. Harding et al. Table 1. Summary of experimental procedures
Group 1 2 3 4
(n=3) (n=3) (n=4) (n=4)
L4 SN section (three days)
Injection of DiY (six days)
Sciatic nerve crush (nine days)
Final experiment (perfusion)
✔ ✔ ✔ ✔
— ✔ ✔ ✔
— — — ✔
9 days 9 days 21 days 21 days
observations indicate that this is indeed the case. Motoneurons can be maintained in a growing state, by nerve injury when the axons attempt to regenerate to replace their lost peripheral stump, or by blocking neuromuscular transmission with á-bungarotoxin, which induces sprouting of axons. Both these interventions have been shown to result in motoneuron death.13,22,27 Conversely, when nerve terminals of immature motoneurons are induced to release transmitter prematurely, their motoneurons become more resistant to target deprivation and many more survive following neonatal nerve injury.17 Therefore, it appears that motoneurons have to undergo a change from a growing into a transmitting cell to survive and mature within the spinal cord.16 In this study, we examined the possibility that maintaining developing nerve terminals in an environment that is conducive to axonal growth, such as that in partially denervated muscle,3 will delay maturation of their motoneurons. In the present experiments, the effect of removing a large proportion of the innervation to the extensor digitorum longus (EDL) and tibialis anterior (TA) muscles on the remaining uninjured motoneurons was examined. In view of previous findings,1 it is likely that this procedure would have prevented the withdrawal of the uninjured axons from the neuromuscular junction, which normally occurs during early postnatal development. We tested whether (i) axons of such motoneurons fail to mature and (ii) their cell bodies remain susceptible to subsequent nerve injury. EXPERIMENTAL PROCEDURES
Four different experimental procedures were carried out on Sprague–Dawley rats (Biological Services, UCL), and these are summarized in Table 1. Surgical procedures Partial denervation. The EDL and TA muscles receive most of their innervation from motor axons that exit the spinal cord in the L4 and L5 ventral roots to form two spinal nerves (SNs). The major neural input to these muscles comes through the L4 SN, which supplies about 80% of the innervation. In three-day-old rats, under halothane anaesthesia (2% in oxygen) and using sterile precautions, the L4 SN was exposed on one side at its exit from the vertebral column and cut. To prevent re-innervation, 2–4 mm of the L4 SN were excised. The L5 SN was left intact. The skin was closed by sutures, and the animals were allowed to recover from the anaesthetic before being returned to their mothers. In all experiments the contralateral unoperated side was
used as a control. The effects of this partial denervation on the ultrastructure of the intact, uninjured axons within the L5 SN were studied in nine-day-old rats. The L5 SN was chosen for examination in preference to the motor nerve to the partially denervated muscles, since the motor nerve contains axons not only from the L5 SN, but also those from the L4 SN which has been injured, and it would be difficult to distinguish the injured axons from the uninjured ones. Fluorescent labelling. Table 1 summarizes the various experimental procedures carried out in this study. In animals of Groups 2, 3 and 4 (see Table 1), motoneurons supplying the EDL and TA muscles were retrogradely labelled with the fluorescent dye Diamidino Yellow (DiY; Illing, Germany). Following partial denervation at three days of age, the rats were re-anaesthetized three days later and under halothane anaesthesia the EDL and TA muscles of both hindlimbs were injected with 2–4 ìl of DiY (2% solution in sterile water) using a fine 5 ìl Hamilton syringe. In this way, all motoneurons that had their axons in the EDL and TA muscles after transection of the L4 SN were labelled with DiY. To ensure that the label reached the motoneuron cell body within three days the animals in Group 2 (see Table 1) were assessed when they were nine days old, and the numbers of labelled motoneurons on the operated and control sides of the spinal cord were counted. In Group 3 (see Table 1), the longer term survival of motoneurons was assessed when the animals were 21 days old. Sciatic nerve crush. Animals in Group 4 (see Table 1) were re-anaesthetized at nine days of age and the sciatic nerve in one hindlimb (on the same side as the partial denervation) was crushed proximal to the division of the tibial and common peroneal nerves, with a pair of fine watchmakers’ forceps. The nerve was then examined to ensure that the epineurial sheath was intact and the nerve translucent. The incision was closed by sutures, and after recovery from the anaesthetic the animals were returned to their mothers. These animals were examined when they were 21 days old. Electron and light microscopy Nine-day-old animals (Group 1) were perfused transcardially under terminal anaesthesia (chloral hydrate, 1 ml/ 100 g body weight, i.p.) with a mixture of 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.3). The L5 SN was dissected from both the operated and contralateral control sides of the rat, postfixed overnight at 4)C and then transferred into 0.1 M phosphate buffer and fixed with 1% OsO4. The nerves were then stained with uranyl acetate (2%), dehydrated and embedded in Araldite. Semi-thin sections (1 ìm) were cut with a glass knife and stained with Toluidine Blue for examination under a light microscope (Zeiss), and adjacent ultra-thin sections (70–90 nm) were then cut with a diamond knife and collected on mesh grids and single slots coated with Formvar plastic film. The sections were counterstained with lead citrate and examined using a transmission electron microscope (Jeol EM1010).
Partial denervation affects intact motoneurons Rats in Groups 2 and 3 were studied either at nine or 21 days. They were perfused transcardially with 4% paraformaldehyde in phosphate buffer, and the spinal cords removed. The spinal cords were postfixed in 4% paraformaldehyde for 2–4 h at 4)C and then transferred into a 30% sucrose solution in phosphate buffer. The lumbar region of the spinal cord was dissected, and serial sections (30 ìm) were cut on a freezing microtome and mounted on to gelatinized slides. The number of retrogradely labelled fluorescent motoneurons within the EDL/TA motor pool was counted under a fluorescent microscope on both the operated and control sides of each spinal cord. The final value was calculated according to Clarke’s Correction Factor by multiplying the raw motoneuron count by an unbiased correction factor, to allow for sectioned particles being counted more than once.7 This correction factor takes motoneuron size into account. RESULTS
293
seen in Fig. 3, which shows a higher magnification of an area of unmyelinated fibres, these fibres were of extraordinarily small diameter for axons from an SN of this age. Furthermore, these axons were not separated by Schwann cell cytoplasm, unlike normal axons in a nerve of this age, and together with the fact that these axons have such a small diameter, this provides further evidence to suggest that these axons were immature. Thus, the uninjured L5 SN on the operated side, part of which supplies the partially denervated EDL and TA muscles, contains immature axons that are not seen in the control L5 SN. These results indicate that removal of a proportion of the innervation to the EDL and TA muscles delayed the maturation of the remaining undamaged axons in the L5 SN.
Ultrastructure of the L5 spinal nerve after removal of the L4 spinal nerve
Motoneuron survival
In rats of Group 1, the L4 SN was sectioned on one side at three days, and six days later the uninjured L5 SN that contained the undamaged axons to the partially denervated EDL and TA muscles, and the contralateral control L5 SN, were removed and processed for ultrastructural analysis (see Table 1). Figure 1 shows typical examples of semi-thin sections (0.5 ìm) of the L5 SN taken from both sides of a nine-day-old rat in which the L4 SN was removed from one side at three days of age. Figure 1A shows most of a cross-section of an L5 SN from the contralateral control side, and illustrates the distribution of myelinated axons, normally present throughout the whole nerve. A higher power view of a small area of this nerve is illustrated in Fig. 1B, showing an even distribution of these myelinated axons. Figure 1C is a light micrograph of a cross-section of the uninjured L5 SN from the operated side in which the L4 SN was sectioned six days earlier. It can be seen that there are large areas of the nerve where no myelinated axons are present. A higher power view from this nerve is shown in Fig. 1D, which illustrates in more detail the absence of myelinated fibres in the area of this nerve. Figure 2A shows a typical example of an electron micrograph of the L5 SN taken from the control side of a nine-day-old animal. It illustrates that the control L5 SN contains a mixture of myelinated and unmyelinated axons. These were intermingled and distributed throughout the cross-section of the nerve. In contrast, the uninjured L5 SN from the operated side where the L4 SN was sectioned six days earlier had areas that contain mainly unmyelinated axons, with a conspicuous absence of myelinated fibres (see Fig. 2B, small arrows). The electron micrograph also shows an adjacent area where myelinated fibres are intermingled with unmyelinated ones, and in these areas of the nerve trunk the myelinated axons were of a similar diameter and the myelin of similar thickness to those in the control L5 SN. In addition, as can be
To establish the number of motoneurons that had their axons in the EDL and TA muscles after removal of the L4 SN, the EDL and TA muscles were injected with the retrograde tracer DiY at six days, and the spinal cords of these rats were examined at nine days (Group 2). On the operated side, 33.1 (&4.0 S.E.M., n=3) motoneurons were labelled and 131.6 (&14.8 S.E.M., n=3) labelled motoneurons were seen on the control side. Thus, the operated side had 26.1% (&5.5% S.E.M., n=3) of the number of labelled motoneurons of the control side (see Table 2). The number of motoneurons innervating the TA and EDL muscles following removal of the L4 SN was then assessed in 21-day-old animals (Group 3). The results show that the number of labelled motoneurons in the spinal cord on the operated side was 34.0 (&4.3 S.E.M., n=4) and that on the control side was 166.7 (&11.4 S.E.M., n=4). When expressed as a percentage of the control side, 20.7% (&3.0% S.E.M., n=4) of motoneurons were present on the operated side of the spinal cord (see Table 2). This is a similar value to that observed in animals examined at nine days of age, and together these results show that (i) the retrograde transport of the fluorescent dye was complete three days after its injection, and (ii) no significant loss of motoneurons in the L5 motor pool has occurred between six and 21 days of age. Whether uninjured motoneurons that innervate the EDL and TA muscles after removal of the L4 SN can survive axon injury was tested next. In rats of Group 4, partial denervation was carried out at three days of age, and three days later the motoneurons to the TA and EDL muscles in both hindlimbs were retrogradely labelled with DiY. When the animals were nine days old, the sciatic nerve was crushed on the operated side. Previous studies have shown that, in normal nine-day-old rats, motoneurons no longer die after sciatic nerve injury.20 However, in the present study, following partial denervation and subsequent nerve injury at nine days of age, the mean number of motoneurons on the operated side of the spinal cord
294
D. I. Harding et al.
Fig. 1. The light micrographs show typical examples of semi-thin sections (0.5 ìm) of the L5 SN taken from a nine-day-old rat on the control unoperated side (A), and on the side where the L4 SN was removed at three days of age (C). In addition, the area enclosed by the dotted lines is shown at higher power from the control side (B), and the partially denervated side (D). A and B illustrate the normal distribution of myelinated axons throughout the whole nerve from the control unoperated side. C and D show that, within the cross-section of the uninjured L5 SN from the operated side where the L4 SN was sectioned six days earlier, there are large areas of the nerve with no myelinated axons present. Scale bars=50 ìm (A, C), 10 ìm (B, D).
Fig. 2. The two electron micrographs show part of a transverse section through the L5 SN from a nine-day-old rat on the control unoperated side (A) and on the side where the L4 SN was sectioned at three days of age (B). In A, large numbers of myelinated axons (large arrows) intermingled with bundles of unmyelinated axons (small arrow) are seen. B shows an area on the right where myelinated (large arrow) and unmyelinated axons (small arrows) are intermingled. This area resembles the control SN seen in A. However, it can be seen that, on the left side of the micrograph, there is a large area that contains only small unmyelinated axons. Such areas were never seen in the control spinal nerves. Scale bar=3 ìm.
Partial denervation affects intact motoneurons
Fig. 2.
295
Fig. 3.
Partial denervation affects intact motoneurons
297
Table 2. The number of motoneurons innervating the tibialis anterior/extensor digitorum longus muscles after partial denervation and sciatic nerve crush Experimental Group 2 3 4
Number of labelled motoneurons&S.E.M. Procedure L4 SN section only; nine days old (n=3) L4 SN section only; 21 days old (n=4) L4 SN section+sciatic nerve crush; 21 days old (n=4)
Control side
Operated side
Operated/control (%)
131.6&14.8 166.7&11.4 170.0&11.8
33.1&4.0 34.0&4.3 8.2&3.4
26.1&5.5 20.7&3.0 4.9&2.0
The numbers of retrogradely labelled motoneurons on the operated side expressed as a percentage of the control in the ventral horns of spinal cords of animals that had their L4 spinal nerve sectioned at three days of age, and DiY injected into their TA and EDL muscles three days later. The values from animals examined at nine days (Group 2) and 21 days (Group 3) of age are shown. The table also shows values obtained from 21-day-old rats that, in addition to the section of the L4 SN, had their sciatic nerve crushed at nine days (Group 4). There is a significantly smaller percentage of labelled motoneurons present in the spinal cords of animals that, in addition to partial denervation, had their sciatic nerve injured (Mann–Whitney U-test, P<0.02). The mean values&S.E.M. are given and n is the number of animals in each group.
was 8.2 (&3.4 S.E.M., n=4) and that on the control side 170.0 (&11.8 S.E.M., n=4). When expressed as a percentage of the control side only 4.9% (&2.0% S.E.M., n=4) of motoneurons survived (see Table 2). This value is significantly smaller than the 20.7% seen after partial denervation only, and shows that most of the motoneurons to the partially denervated TA and EDL muscles were unable to survive nerve injury at nine days of age. These results indicate that developing uninjured motoneurons that have their axons in a partially denervated muscle remain vulnerable to nerve injury for longer than normal, and die as a result of such an insult. DISCUSSION
In the present study, changes of intact motor axons and motoneurons to muscles that had lost most of their innervation were examined. The reduction of innervation was achieved by partial denervation of EDL and TA muscles at three days of age. The results of this study show that removal of the main source of innervation to these muscles at three days of age (i) prevents the maturation of the remaining intact motor axons and (ii) maintains the susceptibility of their motoneurons to axotomy-induced cell death. The possibility that the present results are caused by damage to the L5 SN during the initial operation is unlikely, since we have tested, in this laboratory in a previous study,24 the effect of the removal of the L4 SN on the innervation of the soleus muscle, supplied predominantly by the L5 SN. We found that, after removal of the L4 SN, the soleus muscle develops 85.5% (&5.11% S.E.M., n=6) of the force developed by control muscles, and has 17–25 (n=6) motor units.
This reduction of motor unit number from 30 is consistent with findings that the L4 SN contributes 10–12 motor axons to the innervation of the soleus.5,21 Therefore, it is unlikely that damage to the L5 SN contributes to the present results. It is possible that the changes in the intact motor axons and their motoneurons after partial denervation are due to the response of their nerve terminals to the altered environment in the muscles. During postnatal development of most mammalian muscles, elimination of excess nerve terminals occurs.26 Coincident with this elimination of nerve terminals from the neuromuscular junction, branching of axons in the muscle is arrested. Within the motoneuron cell body, the synthesis of tubulin alpha 1, which is associated with longitudinal growth of axons, is suppressed, and the synthesis of neurofilaments, which are associated with radial growth of axons, is enhanced.19 It has been shown previously that removing a large part of the muscle’s innervation arrests or delays the retraction of the axon terminals of the remaining innervation.1,11,29 Thus, the structural integrity of a larger number of terminals than normal has to be maintained. It could be that under these conditions the reduction in the synthesis of tubulin alpha 1 and increase in the synthesis of neurofilaments in the motoneuron cell body will not occur, since it has been shown that whenever axons grow, for example after injury to adult nerves or during outgrowth of neurites in culture, tubulin levels increase.10,18,25 In addition, if axons are encouraged to sprout during early postnatal development by treating hindlimb muscles of immature mice with botulinum toxin, the downregulation of growth-associated proteins such as growth-associated protein-43 and tubulin alpha 1 fails to occur in motoneurons to these muscles.6 It is possible that partial denervation may affect
Fig. 3. A high-power electron micrograph of an L5 SN from a nine-day-old rat that was partially denervated at three days of age is shown. The micrograph shows the presence of many small-diameter, unmyelinated axons (small arrows). Many of these axons are not separated by intervening Schwann cell cytoplasm (Sc). Scale bar=1 ìm.
298
D. I. Harding et al.
motoneurons in a similar manner. Indeed, an upregulation of growth-associated protein-43 has been reported following partial denervation in adult animals.2 Our present results showing the exceptionally small diameter of the undamaged axons in the L5 SN, some of which are likely to supply the partially denervated muscles, indicate that these axons fail to undergo the radial growth typical of axons to normal muscles. It therefore appears that the retraction of nerve terminals and cessation of their longitudinal growth may be a stimulus for the maturation of the axon and motoneuron. It is interesting in this context that, following nerve crush in neonatal animals where many of the motoneurons that had their axons injured die,22,27,28 the regenerated axons of the surviving motoneurons are extremely small and are unable to attain their normal diameter even after long periods of recovery.31 Although the axons examined in the present study have not been injured, they may have been influenced by events associated with the different conditions of their nerve terminals in the muscle after partial denervation. It is possible that the changes that normally occur in the motoneuron as a consequence of target interaction have failed to take place in motoneurons to partially denervated muscles, so that there is a reduction or a delay in the degree of radial growth of their axons. It is well known that immature motoneurons die when their axons are injured and that this response of motoneurons to axotomy changes dramatically within the first postnatal week of life.14,15,20,22,27,28 This period of development coincides with the time when nerve terminals are no longer elongating, superfluous terminals are retracting and those which remain in contact with the muscle are increasing in volume.9,26 The results of the present study show that intact motoneurons to partially denervated muscles die after injury to their peripheral nerve. It is not clear why these motoneurons die after axotomy when this injury is inflicted relatively late in development. Previous results demonstrate that if, during normal development, the nerve to the TA and EDL muscles is injured in five-day-old rats, motoneurons survive this insult.14,20,27 The results of this study show that this is not the case when the axons supply a muscle that has previously lost part of its innervation. We would like to propose that the difference between the
response of motoneurons to axotomy in these two situations is due to a delay of maturation of axons and motoneurons that have their terminals in a muscle that has been deprived of a large proportion of its innervation. This possibility is supported not only by the fact that these motoneurons remain vulnerable to nerve injury for a prolonged period, but also by the presence of morphologically immature axons in the uninjured L5 SN supplying the partially denervated muscles. This delay in maturation may be due to the failure of the motor nerve terminals to complete the transition from growing to transmitting structures. It is known that, as a result of contact with a muscle cell, the growth of motor nerve terminals is arrested and transmitter release initiated and up-regulated.12,30 Following partial denervation of the EDL and TA muscles at three days of age, some endplates will have lost all or most of their innervation, leaving many muscle fibres denervated. This is likely to lead to a reduction in muscle activity, and it is known that inactive muscles provide an ideal environment for the growth of axons.3,4,8
CONCLUSION
We propose that many of the remaining uninjured axon terminals in these partially denervated muscles, whose axons originate from the L5 SN, will either revert to, or remain, growing structures in order to innervate the otherwise denervated muscle fibres. It is likely that this will delay the transition of their motoneurons from growing into transmitting cells. This prolonged period of axonal growth may account for the fact that these motoneurons remain susceptible to death following nerve injury. Whether these motoneurons that have their axon terminals in partially denervated muscles will remain permanently more susceptible to nerve injury is not clear from the present results. It would be interesting to test this possibility and to relate more precisely the fate of nerve terminals in the muscle to the changes in the motoneuron. Acknowledgements—We are grateful to Action Research, The Wellcome Trust and The Muscular Dystrophy Group of Great Britain. We would also like to thank Mark Turmaine and Jim Dick for their invaluable technical help.
REFERENCES
1.
Betz W. J., Caldwell J. A. and Ribchester R. R. (1980) The effects of partial denervation at birth on the development of muscle fibres and motor units in rat lumbrical muscle. J. Physiol. 303, 265–279. 2. Bisby M. A., Tetzlaff W. and Brown M. C. (1996) GAP-43 mRNA in mouse motoneurons undergoing axonal sprouting in response to muscle paralysis or partial denervation. Eur. J. Neurosci. 8, 1240–1248. 3. Brown M. C., Holland R. L. and Ironton R. (1980) Nodal and terminal sprouting from motor nerves from fast and slow muscles of the mouse. J. Physiol. 306, 493–510. 4. Brown M. C. and Ironton R. (1977) Motoneurone sprouting induced by prolonged tetrodotoxin block of nerve action potentials. Nature 265, 459–461. 5. Brown M. C., Jansen J. K. S. and Van Essen D. (1976) Polyneuronal innervation of skeletal muscle in new-born rats and its elimination during maturation. J. Physiol. 261, 387–442.
Partial denervation affects intact motoneurons 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.
299
Caroni P. and Becker M. (1992) The downregulation of growth-associated proteins in motoneurons at the onset of synapse elimination is controlled by muscle activity and IGF1. J. Neurosci. 12, 3849–3861. Clarke P. G. H. (1993) An unbiased correction factor for cell counts in histological sections. J. Neurosci. Meth. 49, 133–140. Duchen L. W. and Strich S. J. (1968) The effects of botulinum toxin on the pattern of innervation of skeletal muscle in the mouse. Q. Jl exp. Physiol. 53, 84–89. Duxson M. J. (1982) The effect of postsynaptic block on development of the neuromuscular junction in postnatal rats. J. Neurocytol. 11, 395–408. Fine R. E. and Bray D. (1971) Actin in growing nerve cells. Nature New Biol. 234, 115–118. Fisher T. J., Vrbova´ G. and Wijetunge A. (1989) Partial denervation of the rat soleus muscle at two different developmental stages. Neuroscience 28, 755–763. Frank E. and Fischbach G. D. (1979) Early events in neuromuscular junction formation in vitro. Induction of acetylcholine receptor clusters in the postsynaptic membrane and morphology of newly formed synapses. J. Cell Biol. 83, 143–158. Greensmith L. and Vrbova´ G. (1992) Alterations of nerve–muscle interaction during postnatal development influence motoneurone survival in rats. Devl Brain Res. 69, 125–131. Greensmith L., Hasan H. I. and Vrbova´ G. (1994) Nerve injury increases the susceptibility of motoneurons to N-methyl--aspartate-induced neurotoxicity in the developing rat. Neuroscience 58, 727–733. Greensmith L., Mentis G. Z. and Vrbova´ G. (1994) Blockade of N-methyl--aspartate receptors by MK-801 (dizocilpine maleate) rescues motoneurones in developing rats. Devl Brain Res. 81, 162–170. Greensmith L. and Vrbova´ G. (1996) Motoneurone survival: a functional approach. Trends Neurosci. 19, 450–455. Greensmith L., Dick J., Emanuel A. O. and Vrbova´ G. (1996) Induction of transmitter release at the neuromuscular junction prevents motoneuron death after axotomy in neonatal rats. Neuroscience 71, 213–220. Hoffman P. N., Cleveland D. W. and Griffin J. W. (1987) Neurofilament gene expression: a major determinant of axonal caliber. Proc. natn. Acad. Sci. U.S.A. 84, 3472–3476. Hoffman P. N. (1988) Distinct roles of neurofilament and tubulin expression in axonal growth. In Plasticity of the Neuromuscular System (eds Evered D. and Whelan J.), pp. 192–204. Wiley, Chichester. Lowrie M. B., Krishnan S. and Vrbova´ G. (1982) Recovery of slow and fast muscles following nerve injury during early post-natal development in the rat. J. Physiol., Lond. 331, 51–66. Lowrie M. B., O’Brien R. A. D. and Vrbova´ G. (1985) The effect of altered peripheral field on motoneurone function. J. Physiol. 368, 513–524. Lowrie M. B., Subramanian K. and Vrbova´ G. (1987) Permanent changes in muscles and motoneurons induced by nerve injury during a critical period of development of the rat. Devl Brain Res. 310, 91–101. Lowrie M. B. and Vrbova´ G. (1992) Dependence of postnatal motoneurones on their targets: review and hypothesis. Trends Neurosci. 15, 80–84. Maudarbocus S. H. (1987) Responses of slow and fast motor units to partial denervation at different stages in development. B.Sc. thesis, submitted to the University of London, p. 10. Neumann D., Scherson T., Ginzburg I., Littauer U. Z. and Schwartz M. (1983) Regulation of mRNA levels for microtubule proteins during nerve regeneration. Fedn Eur. biochem. Socs Lett. 162, 270–276. O’Brien R. A., O } stberg A. J. and Vrbova´ G. (1978) Observations on the elimination of polyneuronal innervation in developing mammalian skeletal muscle. J. Physiol., Lond. 282, 571–582. Romanes G. J. (1946) Motor localisation and the effects of nerve injury on the ventral horn cells of the spinal cord. J. Anat. 80, 11–131. Schmalbruch H. (1984) Motoneurone death after sciatic nerve section in newborn rats. J. comp. Neurol. 224, 252–258. Thompson W. and Jansen J. K. S. (1977) The extent of sprouting of remaining motor units in partially denervated immature and adult rat soleus muscle. Neuroscience 2, 523–535. Xie Z. and Poo M. (1986) Initial events in the formation of neuromuscular synapse: rapid induction of acetylcholine release from embryonic neuron. J. Cell Biol. 83, 7069–7073. Zelena´ J. and Hnı´k P. (1963) Motor and receptor units in the soleus muscle after nerve regeneration in very young rats. Physiol. Bohemoslov. 12, 227–289. (Accepted 19 January 1998)
PII:
Neuroscience Vol. 86, No. 1, pp. 291–299, 1998 Copyright ? 1998 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/98 $19.00+0.00 S0306-4522(98)00037-2
MOTONEURONS INNERVATING PARTIALLY DENERVATED RAT HINDLIMB MUSCLES REMAIN SUSCEPTIBLE TO AXOTOMY-INDUCED CELL DEATH D. I. HARDING, L. GREENSMITH, P. N. ANDERSON and G. VRBOVA u* Department of Anatomy and Developmental Biology, The Centre for Neuroscience, University College London, Gower Street, London WC1E 6BT, U.K. Abstract––Tibialis anterior and extensor digitorum longus muscles were partially denervated by cutting the L4 spinal nerve in three-day-old rats. The ultrastructure of the intact axons to these muscles in the L5 spinal nerve was examined in nine-day-old rats. In the control L5 spinal nerve, myelinated and unmyelinated axons were intermingled throughout the cross-section of the nerve, while on the operated side the nerve contained areas with predominantly small unmyelinated immature axons. The number of motoneurons innervating the partially denervated muscles was established by retrograde labelling with Diamidino Yellow. In nine- and 21-day-old rats, the number of labelled motoneurons on the partially denervated side, expressed as a percentage of the control side, was 26.1&5.5% and 20.7&3.0%, respectively. The response of these uninjured motoneurons to axotomy was tested. The axons of the motoneurons to the partially denervated muscles were crushed at nine days and the numbers of labelled motoneurons in the spinal cord of these rats counted at 21 days of age. Only 4.9&2.0% labelled motoneurons were seen on the operated side, as opposed to 20.7&3.0% present in animals without sciatic nerve injury. In normal animals, nerve injury at nine days does not cause motoneuron death. Thus, motoneurons to partially denervated muscles (i) have axons with several immature features and (ii) remain susceptible to axotomy-induced death for much longer than normal. ? 1998 IBRO. Published by Elsevier Science Ltd. Key words: axon ultrastructure, motoneuron death, neuronal maturation, target interaction.
Motoneurons of newborn animals die following injury to their axons. In rats, sciatic nerve injury at birth leads to the death of 80–90% of motoneurons, whilst the same insult carried out at five days of age results in little, if any, motoneuron death.20,22 Thus, during the first five days of life motoneurons undergo fundamental changes that enable them to withstand nerve injury. There is some evidence to suggest that this transition, from a motoneuron that is dependent upon contact with its target to one that can survive separation, is induced by functional interaction with the muscle during this critical developmental period. For example, motoneuron death was observed in the spinal cord of rats where active interaction of the motor nerve terminals with the muscle fibres was prevented by paralysing the muscle with á-bungarotoxin during the first few days of postnatal development.13 In these experiments, no injury to the motor axons was inflicted, but nevertheless a substantial loss of motoneurons occurred. These results indicate that some event that is related to the interaction between the nerve terminals and their muscle fibres is involved in the changes that occur in the motoneuron that later enables it to survive target deprivation. *To whom correspondence should be addressed. Abbreviations: DiY, Diamidino Yellow; EDL, extensor digitorum longus; SN, spinal nerve; TA, tibialis anterior. 291
Several proposals have been made as to the precise role of the target in the survival and development of motoneurons, but the exact nature of the target on the developing motoneuron is still not clear. Thus, a better understanding of the nature of the influence of the target on motoneuron survival has yet to be established. The early postnatal period marks a time when motoneurons are not only dependent upon their target for survival, but also a time when both the developing nerve terminals and the postsynaptic membrane are undergoing changes as they mature. One of these changes is the arrest of growth of the nerve terminal and an increase in transmitter release triggered by its encounter with muscle fibres.12,30 This change in the nerve terminal from a growing into a transmitting structure is reflected within the cell body, so that the motoneuron also changes from a growing into a transmitting cell. It has been proposed that the motoneuron must undergo this transition within a specific period of development if it is to mature appropriately and survive within the developing spinal cord.16 According to this hypothesis, motoneurons that are maintained in a growing mode for a prolonged time during their development fail to mature and are unable to survive the effects of increased excitatory inputs within the developing spinal cord.23 Several
292
D. I. Harding et al. Table 1. Summary of experimental procedures
Group 1 2 3 4
(n=3) (n=3) (n=4) (n=4)
L4 SN section (three days)
Injection of DiY (six days)
Sciatic nerve crush (nine days)
Final experiment (perfusion)
✔ ✔ ✔ ✔
— ✔ ✔ ✔
— — — ✔
9 days 9 days 21 days 21 days
observations indicate that this is indeed the case. Motoneurons can be maintained in a growing state, by nerve injury when the axons attempt to regenerate to replace their lost peripheral stump, or by blocking neuromuscular transmission with á-bungarotoxin, which induces sprouting of axons. Both these interventions have been shown to result in motoneuron death.13,22,27 Conversely, when nerve terminals of immature motoneurons are induced to release transmitter prematurely, their motoneurons become more resistant to target deprivation and many more survive following neonatal nerve injury.17 Therefore, it appears that motoneurons have to undergo a change from a growing into a transmitting cell to survive and mature within the spinal cord.16 In this study, we examined the possibility that maintaining developing nerve terminals in an environment that is conducive to axonal growth, such as that in partially denervated muscle,3 will delay maturation of their motoneurons. In the present experiments, the effect of removing a large proportion of the innervation to the extensor digitorum longus (EDL) and tibialis anterior (TA) muscles on the remaining uninjured motoneurons was examined. In view of previous findings,1 it is likely that this procedure would have prevented the withdrawal of the uninjured axons from the neuromuscular junction, which normally occurs during early postnatal development. We tested whether (i) axons of such motoneurons fail to mature and (ii) their cell bodies remain susceptible to subsequent nerve injury. EXPERIMENTAL PROCEDURES
Four different experimental procedures were carried out on Sprague–Dawley rats (Biological Services, UCL), and these are summarized in Table 1. Surgical procedures Partial denervation. The EDL and TA muscles receive most of their innervation from motor axons that exit the spinal cord in the L4 and L5 ventral roots to form two spinal nerves (SNs). The major neural input to these muscles comes through the L4 SN, which supplies about 80% of the innervation. In three-day-old rats, under halothane anaesthesia (2% in oxygen) and using sterile precautions, the L4 SN was exposed on one side at its exit from the vertebral column and cut. To prevent re-innervation, 2–4 mm of the L4 SN were excised. The L5 SN was left intact. The skin was closed by sutures, and the animals were allowed to recover from the anaesthetic before being returned to their mothers. In all experiments the contralateral unoperated side was
used as a control. The effects of this partial denervation on the ultrastructure of the intact, uninjured axons within the L5 SN were studied in nine-day-old rats. The L5 SN was chosen for examination in preference to the motor nerve to the partially denervated muscles, since the motor nerve contains axons not only from the L5 SN, but also those from the L4 SN which has been injured, and it would be difficult to distinguish the injured axons from the uninjured ones. Fluorescent labelling. Table 1 summarizes the various experimental procedures carried out in this study. In animals of Groups 2, 3 and 4 (see Table 1), motoneurons supplying the EDL and TA muscles were retrogradely labelled with the fluorescent dye Diamidino Yellow (DiY; Illing, Germany). Following partial denervation at three days of age, the rats were re-anaesthetized three days later and under halothane anaesthesia the EDL and TA muscles of both hindlimbs were injected with 2–4 ìl of DiY (2% solution in sterile water) using a fine 5 ìl Hamilton syringe. In this way, all motoneurons that had their axons in the EDL and TA muscles after transection of the L4 SN were labelled with DiY. To ensure that the label reached the motoneuron cell body within three days the animals in Group 2 (see Table 1) were assessed when they were nine days old, and the numbers of labelled motoneurons on the operated and control sides of the spinal cord were counted. In Group 3 (see Table 1), the longer term survival of motoneurons was assessed when the animals were 21 days old. Sciatic nerve crush. Animals in Group 4 (see Table 1) were re-anaesthetized at nine days of age and the sciatic nerve in one hindlimb (on the same side as the partial denervation) was crushed proximal to the division of the tibial and common peroneal nerves, with a pair of fine watchmakers’ forceps. The nerve was then examined to ensure that the epineurial sheath was intact and the nerve translucent. The incision was closed by sutures, and after recovery from the anaesthetic the animals were returned to their mothers. These animals were examined when they were 21 days old. Electron and light microscopy Nine-day-old animals (Group 1) were perfused transcardially under terminal anaesthesia (chloral hydrate, 1 ml/ 100 g body weight, i.p.) with a mixture of 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.3). The L5 SN was dissected from both the operated and contralateral control sides of the rat, postfixed overnight at 4)C and then transferred into 0.1 M phosphate buffer and fixed with 1% OsO4. The nerves were then stained with uranyl acetate (2%), dehydrated and embedded in Araldite. Semi-thin sections (1 ìm) were cut with a glass knife and stained with Toluidine Blue for examination under a light microscope (Zeiss), and adjacent ultra-thin sections (70–90 nm) were then cut with a diamond knife and collected on mesh grids and single slots coated with Formvar plastic film. The sections were counterstained with lead citrate and examined using a transmission electron microscope (Jeol EM1010).
Partial denervation affects intact motoneurons Rats in Groups 2 and 3 were studied either at nine or 21 days. They were perfused transcardially with 4% paraformaldehyde in phosphate buffer, and the spinal cords removed. The spinal cords were postfixed in 4% paraformaldehyde for 2–4 h at 4)C and then transferred into a 30% sucrose solution in phosphate buffer. The lumbar region of the spinal cord was dissected, and serial sections (30 ìm) were cut on a freezing microtome and mounted on to gelatinized slides. The number of retrogradely labelled fluorescent motoneurons within the EDL/TA motor pool was counted under a fluorescent microscope on both the operated and control sides of each spinal cord. The final value was calculated according to Clarke’s Correction Factor by multiplying the raw motoneuron count by an unbiased correction factor, to allow for sectioned particles being counted more than once.7 This correction factor takes motoneuron size into account. RESULTS
293
seen in Fig. 3, which shows a higher magnification of an area of unmyelinated fibres, these fibres were of extraordinarily small diameter for axons from an SN of this age. Furthermore, these axons were not separated by Schwann cell cytoplasm, unlike normal axons in a nerve of this age, and together with the fact that these axons have such a small diameter, this provides further evidence to suggest that these axons were immature. Thus, the uninjured L5 SN on the operated side, part of which supplies the partially denervated EDL and TA muscles, contains immature axons that are not seen in the control L5 SN. These results indicate that removal of a proportion of the innervation to the EDL and TA muscles delayed the maturation of the remaining undamaged axons in the L5 SN.
Ultrastructure of the L5 spinal nerve after removal of the L4 spinal nerve
Motoneuron survival
In rats of Group 1, the L4 SN was sectioned on one side at three days, and six days later the uninjured L5 SN that contained the undamaged axons to the partially denervated EDL and TA muscles, and the contralateral control L5 SN, were removed and processed for ultrastructural analysis (see Table 1). Figure 1 shows typical examples of semi-thin sections (0.5 ìm) of the L5 SN taken from both sides of a nine-day-old rat in which the L4 SN was removed from one side at three days of age. Figure 1A shows most of a cross-section of an L5 SN from the contralateral control side, and illustrates the distribution of myelinated axons, normally present throughout the whole nerve. A higher power view of a small area of this nerve is illustrated in Fig. 1B, showing an even distribution of these myelinated axons. Figure 1C is a light micrograph of a cross-section of the uninjured L5 SN from the operated side in which the L4 SN was sectioned six days earlier. It can be seen that there are large areas of the nerve where no myelinated axons are present. A higher power view from this nerve is shown in Fig. 1D, which illustrates in more detail the absence of myelinated fibres in the area of this nerve. Figure 2A shows a typical example of an electron micrograph of the L5 SN taken from the control side of a nine-day-old animal. It illustrates that the control L5 SN contains a mixture of myelinated and unmyelinated axons. These were intermingled and distributed throughout the cross-section of the nerve. In contrast, the uninjured L5 SN from the operated side where the L4 SN was sectioned six days earlier had areas that contain mainly unmyelinated axons, with a conspicuous absence of myelinated fibres (see Fig. 2B, small arrows). The electron micrograph also shows an adjacent area where myelinated fibres are intermingled with unmyelinated ones, and in these areas of the nerve trunk the myelinated axons were of a similar diameter and the myelin of similar thickness to those in the control L5 SN. In addition, as can be
To establish the number of motoneurons that had their axons in the EDL and TA muscles after removal of the L4 SN, the EDL and TA muscles were injected with the retrograde tracer DiY at six days, and the spinal cords of these rats were examined at nine days (Group 2). On the operated side, 33.1 (&4.0 S.E.M., n=3) motoneurons were labelled and 131.6 (&14.8 S.E.M., n=3) labelled motoneurons were seen on the control side. Thus, the operated side had 26.1% (&5.5% S.E.M., n=3) of the number of labelled motoneurons of the control side (see Table 2). The number of motoneurons innervating the TA and EDL muscles following removal of the L4 SN was then assessed in 21-day-old animals (Group 3). The results show that the number of labelled motoneurons in the spinal cord on the operated side was 34.0 (&4.3 S.E.M., n=4) and that on the control side was 166.7 (&11.4 S.E.M., n=4). When expressed as a percentage of the control side, 20.7% (&3.0% S.E.M., n=4) of motoneurons were present on the operated side of the spinal cord (see Table 2). This is a similar value to that observed in animals examined at nine days of age, and together these results show that (i) the retrograde transport of the fluorescent dye was complete three days after its injection, and (ii) no significant loss of motoneurons in the L5 motor pool has occurred between six and 21 days of age. Whether uninjured motoneurons that innervate the EDL and TA muscles after removal of the L4 SN can survive axon injury was tested next. In rats of Group 4, partial denervation was carried out at three days of age, and three days later the motoneurons to the TA and EDL muscles in both hindlimbs were retrogradely labelled with DiY. When the animals were nine days old, the sciatic nerve was crushed on the operated side. Previous studies have shown that, in normal nine-day-old rats, motoneurons no longer die after sciatic nerve injury.20 However, in the present study, following partial denervation and subsequent nerve injury at nine days of age, the mean number of motoneurons on the operated side of the spinal cord
294
D. I. Harding et al.
Fig. 1. The light micrographs show typical examples of semi-thin sections (0.5 ìm) of the L5 SN taken from a nine-day-old rat on the control unoperated side (A), and on the side where the L4 SN was removed at three days of age (C). In addition, the area enclosed by the dotted lines is shown at higher power from the control side (B), and the partially denervated side (D). A and B illustrate the normal distribution of myelinated axons throughout the whole nerve from the control unoperated side. C and D show that, within the cross-section of the uninjured L5 SN from the operated side where the L4 SN was sectioned six days earlier, there are large areas of the nerve with no myelinated axons present. Scale bars=50 ìm (A, C), 10 ìm (B, D).
Fig. 2. The two electron micrographs show part of a transverse section through the L5 SN from a nine-day-old rat on the control unoperated side (A) and on the side where the L4 SN was sectioned at three days of age (B). In A, large numbers of myelinated axons (large arrows) intermingled with bundles of unmyelinated axons (small arrow) are seen. B shows an area on the right where myelinated (large arrow) and unmyelinated axons (small arrows) are intermingled. This area resembles the control SN seen in A. However, it can be seen that, on the left side of the micrograph, there is a large area that contains only small unmyelinated axons. Such areas were never seen in the control spinal nerves. Scale bar=3 ìm.
Partial denervation affects intact motoneurons
Fig. 2.
295
Fig. 3.
Partial denervation affects intact motoneurons
297
Table 2. The number of motoneurons innervating the tibialis anterior/extensor digitorum longus muscles after partial denervation and sciatic nerve crush Experimental Group 2 3 4
Number of labelled motoneurons&S.E.M. Procedure L4 SN section only; nine days old (n=3) L4 SN section only; 21 days old (n=4) L4 SN section+sciatic nerve crush; 21 days old (n=4)
Control side
Operated side
Operated/control (%)
131.6&14.8 166.7&11.4 170.0&11.8
33.1&4.0 34.0&4.3 8.2&3.4
26.1&5.5 20.7&3.0 4.9&2.0
The numbers of retrogradely labelled motoneurons on the operated side expressed as a percentage of the control in the ventral horns of spinal cords of animals that had their L4 spinal nerve sectioned at three days of age, and DiY injected into their TA and EDL muscles three days later. The values from animals examined at nine days (Group 2) and 21 days (Group 3) of age are shown. The table also shows values obtained from 21-day-old rats that, in addition to the section of the L4 SN, had their sciatic nerve crushed at nine days (Group 4). There is a significantly smaller percentage of labelled motoneurons present in the spinal cords of animals that, in addition to partial denervation, had their sciatic nerve injured (Mann–Whitney U-test, P<0.02). The mean values&S.E.M. are given and n is the number of animals in each group.
was 8.2 (&3.4 S.E.M., n=4) and that on the control side 170.0 (&11.8 S.E.M., n=4). When expressed as a percentage of the control side only 4.9% (&2.0% S.E.M., n=4) of motoneurons survived (see Table 2). This value is significantly smaller than the 20.7% seen after partial denervation only, and shows that most of the motoneurons to the partially denervated TA and EDL muscles were unable to survive nerve injury at nine days of age. These results indicate that developing uninjured motoneurons that have their axons in a partially denervated muscle remain vulnerable to nerve injury for longer than normal, and die as a result of such an insult. DISCUSSION
In the present study, changes of intact motor axons and motoneurons to muscles that had lost most of their innervation were examined. The reduction of innervation was achieved by partial denervation of EDL and TA muscles at three days of age. The results of this study show that removal of the main source of innervation to these muscles at three days of age (i) prevents the maturation of the remaining intact motor axons and (ii) maintains the susceptibility of their motoneurons to axotomy-induced cell death. The possibility that the present results are caused by damage to the L5 SN during the initial operation is unlikely, since we have tested, in this laboratory in a previous study,24 the effect of the removal of the L4 SN on the innervation of the soleus muscle, supplied predominantly by the L5 SN. We found that, after removal of the L4 SN, the soleus muscle develops 85.5% (&5.11% S.E.M., n=6) of the force developed by control muscles, and has 17–25 (n=6) motor units.
This reduction of motor unit number from 30 is consistent with findings that the L4 SN contributes 10–12 motor axons to the innervation of the soleus.5,21 Therefore, it is unlikely that damage to the L5 SN contributes to the present results. It is possible that the changes in the intact motor axons and their motoneurons after partial denervation are due to the response of their nerve terminals to the altered environment in the muscles. During postnatal development of most mammalian muscles, elimination of excess nerve terminals occurs.26 Coincident with this elimination of nerve terminals from the neuromuscular junction, branching of axons in the muscle is arrested. Within the motoneuron cell body, the synthesis of tubulin alpha 1, which is associated with longitudinal growth of axons, is suppressed, and the synthesis of neurofilaments, which are associated with radial growth of axons, is enhanced.19 It has been shown previously that removing a large part of the muscle’s innervation arrests or delays the retraction of the axon terminals of the remaining innervation.1,11,29 Thus, the structural integrity of a larger number of terminals than normal has to be maintained. It could be that under these conditions the reduction in the synthesis of tubulin alpha 1 and increase in the synthesis of neurofilaments in the motoneuron cell body will not occur, since it has been shown that whenever axons grow, for example after injury to adult nerves or during outgrowth of neurites in culture, tubulin levels increase.10,18,25 In addition, if axons are encouraged to sprout during early postnatal development by treating hindlimb muscles of immature mice with botulinum toxin, the downregulation of growth-associated proteins such as growth-associated protein-43 and tubulin alpha 1 fails to occur in motoneurons to these muscles.6 It is possible that partial denervation may affect
Fig. 3. A high-power electron micrograph of an L5 SN from a nine-day-old rat that was partially denervated at three days of age is shown. The micrograph shows the presence of many small-diameter, unmyelinated axons (small arrows). Many of these axons are not separated by intervening Schwann cell cytoplasm (Sc). Scale bar=1 ìm.
298
D. I. Harding et al.
motoneurons in a similar manner. Indeed, an upregulation of growth-associated protein-43 has been reported following partial denervation in adult animals.2 Our present results showing the exceptionally small diameter of the undamaged axons in the L5 SN, some of which are likely to supply the partially denervated muscles, indicate that these axons fail to undergo the radial growth typical of axons to normal muscles. It therefore appears that the retraction of nerve terminals and cessation of their longitudinal growth may be a stimulus for the maturation of the axon and motoneuron. It is interesting in this context that, following nerve crush in neonatal animals where many of the motoneurons that had their axons injured die,22,27,28 the regenerated axons of the surviving motoneurons are extremely small and are unable to attain their normal diameter even after long periods of recovery.31 Although the axons examined in the present study have not been injured, they may have been influenced by events associated with the different conditions of their nerve terminals in the muscle after partial denervation. It is possible that the changes that normally occur in the motoneuron as a consequence of target interaction have failed to take place in motoneurons to partially denervated muscles, so that there is a reduction or a delay in the degree of radial growth of their axons. It is well known that immature motoneurons die when their axons are injured and that this response of motoneurons to axotomy changes dramatically within the first postnatal week of life.14,15,20,22,27,28 This period of development coincides with the time when nerve terminals are no longer elongating, superfluous terminals are retracting and those which remain in contact with the muscle are increasing in volume.9,26 The results of the present study show that intact motoneurons to partially denervated muscles die after injury to their peripheral nerve. It is not clear why these motoneurons die after axotomy when this injury is inflicted relatively late in development. Previous results demonstrate that if, during normal development, the nerve to the TA and EDL muscles is injured in five-day-old rats, motoneurons survive this insult.14,20,27 The results of this study show that this is not the case when the axons supply a muscle that has previously lost part of its innervation. We would like to propose that the difference between the
response of motoneurons to axotomy in these two situations is due to a delay of maturation of axons and motoneurons that have their terminals in a muscle that has been deprived of a large proportion of its innervation. This possibility is supported not only by the fact that these motoneurons remain vulnerable to nerve injury for a prolonged period, but also by the presence of morphologically immature axons in the uninjured L5 SN supplying the partially denervated muscles. This delay in maturation may be due to the failure of the motor nerve terminals to complete the transition from growing to transmitting structures. It is known that, as a result of contact with a muscle cell, the growth of motor nerve terminals is arrested and transmitter release initiated and up-regulated.12,30 Following partial denervation of the EDL and TA muscles at three days of age, some endplates will have lost all or most of their innervation, leaving many muscle fibres denervated. This is likely to lead to a reduction in muscle activity, and it is known that inactive muscles provide an ideal environment for the growth of axons.3,4,8
CONCLUSION
We propose that many of the remaining uninjured axon terminals in these partially denervated muscles, whose axons originate from the L5 SN, will either revert to, or remain, growing structures in order to innervate the otherwise denervated muscle fibres. It is likely that this will delay the transition of their motoneurons from growing into transmitting cells. This prolonged period of axonal growth may account for the fact that these motoneurons remain susceptible to death following nerve injury. Whether these motoneurons that have their axon terminals in partially denervated muscles will remain permanently more susceptible to nerve injury is not clear from the present results. It would be interesting to test this possibility and to relate more precisely the fate of nerve terminals in the muscle to the changes in the motoneuron. Acknowledgements—We are grateful to Action Research, The Wellcome Trust and The Muscular Dystrophy Group of Great Britain. We would also like to thank Mark Turmaine and Jim Dick for their invaluable technical help.
REFERENCES
1.
Betz W. J., Caldwell J. A. and Ribchester R. R. (1980) The effects of partial denervation at birth on the development of muscle fibres and motor units in rat lumbrical muscle. J. Physiol. 303, 265–279. 2. Bisby M. A., Tetzlaff W. and Brown M. C. (1996) GAP-43 mRNA in mouse motoneurons undergoing axonal sprouting in response to muscle paralysis or partial denervation. Eur. J. Neurosci. 8, 1240–1248. 3. Brown M. C., Holland R. L. and Ironton R. (1980) Nodal and terminal sprouting from motor nerves from fast and slow muscles of the mouse. J. Physiol. 306, 493–510. 4. Brown M. C. and Ironton R. (1977) Motoneurone sprouting induced by prolonged tetrodotoxin block of nerve action potentials. Nature 265, 459–461. 5. Brown M. C., Jansen J. K. S. and Van Essen D. (1976) Polyneuronal innervation of skeletal muscle in new-born rats and its elimination during maturation. J. Physiol. 261, 387–442.
Partial denervation affects intact motoneurons 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.
299
Caroni P. and Becker M. (1992) The downregulation of growth-associated proteins in motoneurons at the onset of synapse elimination is controlled by muscle activity and IGF1. J. Neurosci. 12, 3849–3861. Clarke P. G. H. (1993) An unbiased correction factor for cell counts in histological sections. J. Neurosci. Meth. 49, 133–140. Duchen L. W. and Strich S. J. (1968) The effects of botulinum toxin on the pattern of innervation of skeletal muscle in the mouse. Q. Jl exp. Physiol. 53, 84–89. Duxson M. J. (1982) The effect of postsynaptic block on development of the neuromuscular junction in postnatal rats. J. Neurocytol. 11, 395–408. Fine R. E. and Bray D. (1971) Actin in growing nerve cells. Nature New Biol. 234, 115–118. Fisher T. J., Vrbova´ G. and Wijetunge A. (1989) Partial denervation of the rat soleus muscle at two different developmental stages. Neuroscience 28, 755–763. Frank E. and Fischbach G. D. (1979) Early events in neuromuscular junction formation in vitro. Induction of acetylcholine receptor clusters in the postsynaptic membrane and morphology of newly formed synapses. J. Cell Biol. 83, 143–158. Greensmith L. and Vrbova´ G. (1992) Alterations of nerve–muscle interaction during postnatal development influence motoneurone survival in rats. Devl Brain Res. 69, 125–131. Greensmith L., Hasan H. I. and Vrbova´ G. (1994) Nerve injury increases the susceptibility of motoneurons to N-methyl--aspartate-induced neurotoxicity in the developing rat. Neuroscience 58, 727–733. Greensmith L., Mentis G. Z. and Vrbova´ G. (1994) Blockade of N-methyl--aspartate receptors by MK-801 (dizocilpine maleate) rescues motoneurones in developing rats. Devl Brain Res. 81, 162–170. Greensmith L. and Vrbova´ G. (1996) Motoneurone survival: a functional approach. Trends Neurosci. 19, 450–455. Greensmith L., Dick J., Emanuel A. O. and Vrbova´ G. (1996) Induction of transmitter release at the neuromuscular junction prevents motoneuron death after axotomy in neonatal rats. Neuroscience 71, 213–220. Hoffman P. N., Cleveland D. W. and Griffin J. W. (1987) Neurofilament gene expression: a major determinant of axonal caliber. Proc. natn. Acad. Sci. U.S.A. 84, 3472–3476. Hoffman P. N. (1988) Distinct roles of neurofilament and tubulin expression in axonal growth. In Plasticity of the Neuromuscular System (eds Evered D. and Whelan J.), pp. 192–204. Wiley, Chichester. Lowrie M. B., Krishnan S. and Vrbova´ G. (1982) Recovery of slow and fast muscles following nerve injury during early post-natal development in the rat. J. Physiol., Lond. 331, 51–66. Lowrie M. B., O’Brien R. A. D. and Vrbova´ G. (1985) The effect of altered peripheral field on motoneurone function. J. Physiol. 368, 513–524. Lowrie M. B., Subramanian K. and Vrbova´ G. (1987) Permanent changes in muscles and motoneurons induced by nerve injury during a critical period of development of the rat. Devl Brain Res. 310, 91–101. Lowrie M. B. and Vrbova´ G. (1992) Dependence of postnatal motoneurones on their targets: review and hypothesis. Trends Neurosci. 15, 80–84. Maudarbocus S. H. (1987) Responses of slow and fast motor units to partial denervation at different stages in development. B.Sc. thesis, submitted to the University of London, p. 10. Neumann D., Scherson T., Ginzburg I., Littauer U. Z. and Schwartz M. (1983) Regulation of mRNA levels for microtubule proteins during nerve regeneration. Fedn Eur. biochem. Socs Lett. 162, 270–276. O’Brien R. A., O } stberg A. J. and Vrbova´ G. (1978) Observations on the elimination of polyneuronal innervation in developing mammalian skeletal muscle. J. Physiol., Lond. 282, 571–582. Romanes G. J. (1946) Motor localisation and the effects of nerve injury on the ventral horn cells of the spinal cord. J. Anat. 80, 11–131. Schmalbruch H. (1984) Motoneurone death after sciatic nerve section in newborn rats. J. comp. Neurol. 224, 252–258. Thompson W. and Jansen J. K. S. (1977) The extent of sprouting of remaining motor units in partially denervated immature and adult rat soleus muscle. Neuroscience 2, 523–535. Xie Z. and Poo M. (1986) Initial events in the formation of neuromuscular synapse: rapid induction of acetylcholine release from embryonic neuron. J. Cell Biol. 83, 7069–7073. Zelena´ J. and Hnı´k P. (1963) Motor and receptor units in the soleus muscle after nerve regeneration in very young rats. Physiol. Bohemoslov. 12, 227–289. (Accepted 19 January 1998)