283
Brain Research Reviews, 17 (1992) 283-289 0 1992 Elsevier Science Publishers B.V. Ah rights reserved 0165-0173/92/$05.00
BRESR 95000
Short Review
Activity-dependent
development of spinal cord motor neurons
Robert G. Kalb a and Susan Hockfield b n Department of Neurology and b Section of Neurobiology, Yale University School of Medicine, New Haven, CT 04510 (USA) (Accepted 14 July 1992)
Key words: Motor neuron; Spinal cord; Activity; N-Methyl-D-aspartate
receptor; Cat-301; Proteoglycan; Critical period
CONTENTS 1. Introduction..
283
................................................................................
2. Molecular correlates of activity-dependent
development
3. Molecular and anatomical evidence for activity-dependent 4. Involvement of the N-methyl-D-aspartate
.................................................. development of motor neurons
..........................
subtype of glutamate receptor in motor neuron development .................
284
.
285
.
286
5. Patterns of afferent activity can influence physiological properties of motor neurons through activation of the NMDA receptor 6. Concludingremarks
287 287
............................................................................
7. Summary ....................................................................................
288
Acknowledgements
...................................................................................
288
.........................................................................................
288
References
Noteaddedinproof
..................................................................................
1. INTRODUCTION
During the development of the mammalian central nervous system (CNS), the mature set of synaptic connections is generated by a variety of mechanisms over an extended period of time. The choice of initial synaptic sites and partners is thought to be governed by intrinsic cellular properties including trophic and other chemical cues. In many systems, this initial set of synaptic connections is exuberant and is refined to its mature complement, at least in part, by neuronal activity32. Activity-dependent developmental processes have been demonstrated to have a profound effect not only on synaptic connectivity, but also on the morphological and electrophysiological properties of neurons4.36. Ac-
289
tivity-dependent developmental processes take place largely during a circumscribed period in early postnatal life known as the ‘critical’ or ‘sensitive’ period. Understanding the characteristics and mechanisms of these early plastic processes and the subsequent loss of plasticity may have important implications not only for development, but also for the recovery of function after damage to the adult nervous system. The visual system of vertebrates has provided some of the most compelling demonstrations of activity-dependent development. Visually evoked neuronal activity in the early postnatal period has profound effects on the anatomy and physiology of neurons in the visual system of mammals and amphibia. Visually evoked neuronal activity controls the segregation of geniculo-
CorresPondence to: R. Kalb,Department of Neurology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA. Fax: (1) (203) 7855694.
284 cortical input representing right and left eyes into ocular dominance columns in the primary visual cortex of cats36,“8 the segregation of retinal inputs in the lateral geniculate nucleus (LGNJ6, the retinotopic mapping of afferents from the nucleus isthmi onto the tectum in two-eyed frogs4” and the segregation of retinal afferents into ocular dominance stripes in the tectum of three-eyed frogs3’. In all these systems, competition between two inputs with non-synchronous patterns of activity leads to the stabilization of a subset of connections4936. The cellular and molecular biological events underlying the selection and stabilization of mature synaptic organization are just beginning to be explored. Study of developing motor neurons has yielded conflicting results concerning the importance of activity in the acquisition of mature neuronal properties. Despite the profound changes in motor neuron structure that occur during the postnatal period3,34 few studies have described an activity-dependent critical period in motor neuron development. The stereotypical development of many motor neuron characteristics, such as the formation of initial connections between homonomous muscle afferents and their motor neurons’” and the formation of pattern generators for fictive swimming in is thought to be mediated by intrinsic amphibiansI signals and chemical cues. In the chicken, blockade of muscle activity by d-tubocurare application does not perturb the accurate mapping of primary afferents onto motor neurons, although the Ia afferent-induced motor neuron EPSP’s are twice normal sizez9. While these observations suggest that the initial matching of la afferents onto motor neurons is likely to be activityindependent, it does not preclude the existence of activity-dependent sculpting of the synaptic input to motor neurons during the postnatal period. Other characteristics of motor neurons clearly are not hard wired and are likely to be under the control of epigenetic factors. The hormonal control of synapse elimination from sexually dimorphic muscles is an example of a developmental phenomenon controlled by external influences during a critical period in early postnatal life 19-20 . Synapse elimination from the levator ani (LA) muscle in male rats occurs during the first month of postnatal life and is sensitive to manipulation of gonadal androgen levels. Removal of endogenous androgens in juvenile males by castration accelerates the loss of polyinnervation2’. Precocious exposure to elevated levels of androgens prevents the loss of polyinnervation and this polyinnervation persists for months after the cessation of exogenous androgen administration”. These results suggest that physiological levels of androgens during a critical period in early life
normally regulates the maturation of the LA neuromuscular system. Neuronal activity also has been demonstrated to influence the development of other motor neuron characteristics26-28. Synapses at the crustacean neuromuscular junction can be physiologically categorized as tonic or phasic based on neurotransmitter release characteristics and fatigue resistance. Tonic-patterned stimulation of single phasic motor neurons can transform the physiology of motor neurons from phase into more tonic-like. Such phasic-to-tonic transformation is accompanied by the acquisition of tonic morphology (larger terminal arbors with more varicosities) in phasic motor neurons 28. This activity-induced motor neuron plasticity is age-dependent, as its induction is significantly greater in young crayfish than in adults2’. These observations show that the pattern of activity of motor neurons in early life can lead to alterations in motor neuron properties. In the mammalian neuromuscular system, it has recently been demonstrated that patterns of activity (phasic versus tonic) can similarly influence the morphology of mouse DRG neurons in vitro’. Finally, although activity clearly has the potential to influence motor neuron characteristics, it is possible that not all motor neurons are equally susceptible to the effects of activity. In vivo imaging techniques have permitted repeated observations of the mouse neuromuscular junction over extended periods of time2s,43*44. Such studies suggest significant synaptic remodeling of the neuromuscular junction in some but not all adult muscles. The remodelling observed may be activity-dependent, since remodelling was more pronounced in slow-twitch (tonic) muscle fibers than in fast-twitch (phasic) fibers. These differences in the plasticity of muscle types in adults may reflect lifelong differences between motor neurons and their targets. Some types of motor neurons may be embryonically hard-wired while others may be more sensitive to activity-dependent structural and functional modifications. 2. MOLECULAR CORRELATES OF ACTIVI-N-DEPENDENT DEVELOPMENT Our work over the last several years has been based on the hypothesis that changes in, or the acquisition of new, molecular properties might accompany the acquisition of mature synaptic structure. In the cat visual system we have demonstrated that the chondroitin sulfate proteoglycan recognized by monoclonal antibody Cat-3014s is expressed late in the postnatal period and is regulated by neuronal activity during early postnatal life39. In the cat LGN, the cell-surface Cat-301 proteoglycan is expressed on Y-cells in animals raised
285 in a normal visual environment “, appearing between postnatal days 30 and 90 (ref. 39). In animals deprived of normal visual experience by dark rearing13 monocular lid suture39, or intraocular administration of tetrodotoxin’6, the expression of the Cat-301 proteoglycan is dramatically reduced. Visual deprivation in adult animals for extended periods of time has no effect on expression of the proteoglycan. The Cat-301 antigen is, therefore, a unique positive molecular marker of normal activity-dependent development in the cat visual system. 3. MOLECULAR AND ANATOMICAL EVIDENCE FOR ACTIVITY-DEPENDENT
DEVELOPMENT OF MOTOR
NEURONS
We have used Cat-301 to explore the possibility that activity-dependent development might also be a feature of spinal motor neurons. Cat-301 immunohistochemistry combined with retrograde labelling from the sciatic nerve demonstrated that in normal animals all sciatic motor neurons express the Cat-301 antigen*l. In the spinal cord, like in the LGN, Cat-301 immunoreactivity develops relatively late in the postnatal period. Immunoreactivity is not present at birth and
adult levels of Cat-301 expression are not reached until P14 (that is, all sciatic motor neurons are Cat-301 positive at P14). Two other motor neuron antigens (recognized by monoclonal antibodies Rat-302 and Rat-303) are also first expressed during the postnatal period; they, too, are not present at birth and develop to adult levels during the second postnatal week of life*l. Regulation of Cat-301 expression on motor neurons mimics the activity-dependent expression demonstrated in the cat LGN. Connectivity and activity of motor neurons was disrupted in our initial experiments by a quite severe lesion, sciatic nerve crush*l (see Fig. 1). The effect of nerve crush performed at P5 (several days before normal Cat-301 expression) was compared to the effect of nerve crush in adult animals. In animals with sciatic nerve crush at P5, Cat-301 expression was profoundly inhibited: less than 5% of sciatic motor neurons expressed the Cat-301 antigen at P21 (Fig. 1). Importantly, this reduction in Cat-301 expression was relatively specific in that the antigens recognized by Rat-302 and Rat-303 showed no reduction in expression ‘l. Sciatic nerve lesions in adult animals had no effect on Cat-301 expression, demonstrating that antigen expression is not mediated simply by continuous
Fig. 1. Two weeks after neonatal sciatic nerve lesion at P7, retrogradely labeled motor neurons do not express Cat-301 immunoreactivity. A: one motor neuron in this field shows retrograde labeling with Fast blue (arrowhead), 2 weeks after a sciatic nerve lesion. The UV illumination faintly excites the FITC labeling of a Cat-301 positive, Fast blue negative cell in the lower portion of the panel (arrow). Under UV illumination this cell fluoresces a faint yellow-green. B: same field shown in (A) viewed under FITC optics. No surface-associated Cat-301 immunoreactivity (visualized with FITC-conjugated secondary antibody) is seen at the location of the Fast blue cell (arrowhead). As this photomicrograph has been overexposed to more clearly demonstrate the lack of Cat-301 immunoreactivity, the fine intracellular particulate Fast blue staining is seen breaking though under FITC illumination. A Fast blue negative, Cat-301 positive cell is seen in the lower portion of the field (arrow). x390. Taken from Ref. 21.
286 4. INVOLVEMENT
neuronal activity, but specifically by activity during a critical period in development. Subsequent experiments demonstrated that the effect of sciatic nerve crush on Cat-301 expression was not due to direct injury to the motor neuron itself, but was more likely due to a disruption in the pattern of activity of the motor neuron. Lesions of descending input to motor neurons 2’ (by thoracic hemicordotomy) or of primary afferent input22 (by dorsal rhizotomy) also produced reductions in Cat-301 immunoreactivity when performed on early postnatal animals. Thus, lesions that altered input to motor neurons, without directly injuring the motor neurons, decreased Cat-301 expression (Fig. 2). These results suggest that the coordinated input from upper motor neurons and segmental afferents during early postnatal life is required for the normal molecular development of motor neurons. These results parallel observations in the visual system where patterns of neuronal activity in early postnatal life determine the course of neuronal maturation. One set of experiments was designed to determine whether a particular class of primary afferent input was required for Cat-301 expression. Capsaicin, a neurotoxin selective for small diameter afferents, administered to neonates had no effect on Cat-301 expression22, providing evidence that expression of the Cat301 proteoglycan on motor neurons depends upon large diameter primary afferents. Many of these fibers are glutamatergic35.
Thoracic
One class of glutamate receptor, the NMDA receptor, has been linked to developmentally regulated synaptic plasticity in the visual system of cat’,24 and frog 2,4. We have tested whether the NMDA receptor might play a role in the development of motor neurons. Administration of the NMDA antagonists APV or MK-801 during the neonatal period inhibited the expression of the Cat-301 proteoglycan on motor neurons23 (summarized in Fig. 2). This inhibition was dose-dependent, stereospecific and selective for the Cat-301 proteoglycan. The inhibition was also temporally restricted, in that in adult animals Cat-301 immunoreactivity on motor neurons was not diminished by NMDA receptor blockade. The effect of NMDA receptor blockade on motor neuron development is most likely to be mediated by receptors at the segmental level because the efficacy of antagonists decreased in rough proportion to the distance from the sciatic motor neuron pool at which they were administered2’. These results suggest that motor neurons undergo an activity-dependent, NMDA-mediated critical period in development. The role of NMDA receptor in development of the mammalian visual system currently is a topic of debate. In the developing ferret LGN (M. Sur, personal com-
Hemlcordotomy
Capsalcln
Sciatic antagonlsm
N-METHYL-D-ASPAR-
TOR NEURON DEVELOPMENT
Dorsal Rhlzotomy
NMDA receptor
OF THE
TATE SUBTYPE OF GLUTAMATE RECEPTOR IN MO-
- 49% - 45%
- 98%
Nerve Crush - 4%
- 22%
Fig. 2. Influence of neonatal lesions on motor neuron Cat-301 expression. Surgical and pharmacologic lesions of different components of the neuromuscular system was performed on postnatal day 7 (P7) and at P21 Cat-301 immunoreactivity on sciatic motor neurons was assessed. The percentages reported represent the average number of Cat-301 positive motor neurons after each manipulation. Each manipulation was performed on at least three separate animals from different litters and the Cat-301 status was derived by assaying a minimum of 300 retrogradely labeled sciatic motor neurons. Of note, in parallel with the neonatal lesion experiments, adult animals were subjected to the same interventions and the Cat-301 status of sciatic motor neurons was assayed 2 weeks later. When any of the above lesions were performed on adult animals, in all instances, virtually all motor neurons retained Cat-301 immunoreactivity.
287 munication) and cat visual cortex3’ a large component (and perhaps the majority) of the excitatory drive to neurons is mediated by activation of the NMDA receptor, so that NMDA receptor blockade may be equivalent to a non-specific inhibition of neuronal activity’. In contrast to the visual system, experiments on spinal cord neurons suggest that the major excitatory drive to motorneurons during development is not mediated by activation of the NMDA receptor and that a substantial amount of excitation to motorneurons is carried by non-NMDA glutamate receptors in the pre- and early . We have demonstrated that post-natal period 18*46 spinal cord motor neurons can be driven by segmental reflex pathways in the presence of NMDA receptor antagonists I5. NMDA antagonists do not, therefore, globally prevent synaptic activation of motor neurons. Together these results suggest that coordinated action of primary afferents and descending inputs can activate the NMDA receptor at the segmental level (see Note added in proof). Activation of NMDA receptors on target cells may then regulate the expression of a class of neuronal proteins, of which Cat-301 is one example, that might subserve the anatomical and physiological consequences of activity-dependent development. 5. PATTERNS OF AFFERENT ACTIVITY CAN INFLUENCE PHYSIOLOGIC PROPERTIES OF MOTOR NEURONS THROUGH ACTIVATION
OF THE NMDA RE-
CEPTOR
Our in vivo studies provide molecular evidence that normal motor neuron development requires the functioning of NMDA receptors at the spinal segmental level. These observations complement in vitro observations of NMDA receptor mediated changes in synaptic efficacy of ventral horn neurons8Y31. Dissociated cells from the spinal ventral horn (VII) were cultured in the central chamber of a three compartment tissue culture dish while dorsal root ganglion (DRG) neurons were cultured in the two flanking chambers and allowed to grow axons into the central chamber. One of the DRG chambers was chronically stimulated at a level sufficient to drive the VII neurons. This led to an enhanced synaptic efficacy and a larger number of inputs from the stimulated cells in comparison to the inputs from unstimulated control side. The increase in efficacy could be prevented by application of NMDA receptor antagonists during the stimulation period. Further work has shown that this process is likely to be mediated by a calcium conductance’. These experiments demonstrate that the DRG-VH synapses possess the potential for activity-dependent modification. They suggest that
the developing spinal cord, like the developing visual system, may utilize activity-dependent mechanisms to establishing synaptic connectivity.
6. CONCLUDING REMARKS
Studies of neuromuscular system development in many species have shown that extensive electrophysiologica1’2T41and morphological5 changes occur during the early postnatal period. The studies described here provide support for the possibility that the maturation of motor neurons, like neurons in many other systems, has an activity-dependent component. Whether other motor neuron characteristics might be regulated by neuronal activity during development remains to be explored in detail. In the LGN”*‘3,42 and the spinal nucleus of Clarke37, cell soma size and dendritic morphology are altered by activity-deprivation during a critical period in early development. In our ongoing experiments, measurements of retrogradely labeled motor neurons in normal and in animals deprived of normal neuromuscular activity during the neonatal period indicate that activity-deprived motor neurons are significantly smaller in cross-sectional area than control motor neurons. Since cell size and dendritic architecture frequently co-vary, it will be interesting to see if the dendritic arbors of motor neurons are also regulated by neuronal activity during a critical period in early postnatal life. Why motor neurons might have an activity-dependent period in development remains open to speculation at present. Based on insights gleaned from the visual system, one possibility is that activity-dependent processes are required to fine tune the inputs (or their gain) onto motor neurons for optimal neuromuscular performance. A child’s walking, running and jumping might provide a pattern of activation of primary afferent and descending inputs onto motor neurons (and/or a set of interneurons) that could lead to activation of NMDA receptors and activity-dependent modifications of the motor neuron. One might then expect that the activity-driven shaping of the inputs to and response properties of motor neurons would be reflected in dendritic architecture as well as synaptology. The activity-dependent fine tuning of the neuromuscular system might account, in part, individual differences in motor performance. One implication of the existence of a critical period in motor neuron maturation is that the structural and functional adaptability of motor neurons to changing demands has finite temporal limits, which may bear on the functional outcome after injury to the motor system.
288 7. SUMMARY Patterned neuronal activity in early postnatal life can regulate the acquisition of the mature morphological and electrophysiological properties of neurons. Many properties of motor neurons are developmentally regulated and may be influenced by epigenetic factors. The pattern of activation of motor neurons can regulate axon terminal morphology and synaptic efficacy at the neuromuscular junction. Motor neuron morphology and synaptic connections can also be modified by exposure to specific hormones in the early postnatal period. The acquisition of mature physiological and anatomical properties is paralleled by the acquisition of specific molecular properties. Recent experiments using molecular markers for motor neuron differentiation indicate that motor neurons undergo activity-dependent development during a circumscribed period in early postnatal life. Normal motor neuron differentiation requires a normal pattern of neuronal activity in early postnatal life. Differentiation also requires activation of the NMDA receptor over the same time period. The activity-dependent development of morphological, electrophysiological and molecular properties of motor neurons is similar to activity-dependent development in the vertebrate visual system. The neuromuscular system may provide an accessible system for characterizing the molecules subserving the translation of patterned neuronal activity into mature neuronal phenotype. Acknowledgements. This work was supported by Grants to R.G.K. from NIH (NS012471, the Muscular Dystrophy Association and the Daniel Heumann Fund for Spinal Cord Research and to S.H. from NIH (EY-06511) and NSF (BNS-8544681).
REFERENCES 1 Bear, M.F., Kleinschmidt, A., Gu, Q. and Singer, W., Disruption of experience-dependent synaptic modifications in striate cortex by infusion of an NMDA receptor antagonist, J. Neurosci., 10 (1990) 909-925. 2 Cline, H.T., Debski, E.A. and Constantine-Paton, M., N-methylD-aspartate receptor antagonist desegregates eye-specific columns, PNAS, 84 (1987) 4342-4345. 3 Conradi, S. and Ronnevi, L.-O., Ultrastructure and synaptology of initial axon segment of cat spinal motor neurons during early postnatal development, J. Neurocytol., 6 (1977) 195-210. 4 Constantine-Paton, M., Cline, H.T. and Debski, E., Patterned activity, synaptic convergence and the NMDA receptor in developing visual pathways, Annu. Reu. Neurosci., 13 (1990) 129-154. 5 Cummings, J.P. and Stelzner, D.J., Prenatal and postnatal development of lamina IX neurons in the rat thoracic spinal cord, fin. Neural., 83 (1984) 155-166. 6 Dubin, M.W., Stark, L.A. and Archer, S.M., A role for actionpotential activity in the development of neuronal connections in the kitten retinogeniculate pathway, J. Neurosci., 6 (1986) 102l1036. 7 Fields, R.D., Neale, E.A. and Nelson, P.G., Effects of patterned electrical activity on neurite outgrowth from mouse sensory neurons, J. Neurosci., 10 (1990) 2950-2964.
8 Fields, R.D., Yu, C. and Nelson, P.G., Calcium, network activity 9 10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
and the role of NMDA channels in synaptic plasticity in vitro, .I. Neurosci., 11 (1991) 134-146. Frank, E., The influence of neuronal activity on patterns of synaptic connections, TINS, 10 (1987) 188-190. Frank, E. and Jackson, P.C., Normal electrical activity is not required for the formation of specific sensory-motor synapses, Bruin Res., 378 (1986) 147-151. Friedlander, M.J., Stanford, L.R. and Sherman, S.M., Effects of monocular deprivation on the structure-function relationship of individual neurons in the cat’s lateral geniculate nucleus, J. Neurosci., 2 (1982) 321-330. Fulton, B.P. and Walton, K., Electrophysiological properties of neonatal rat motoneurons studied in vitro, J. PhysioL, 370 (1986) 651-678. Guimaraes, A., Zaremba, S. and Hockfield, S., Molecular and morphological changes in cat lateral geniculate nucleus and visual cortex induced by visual deprivation are revealed by monoclonal antibodies Cat-301 and Cat-304, J. Neurosci., 10 (1990) 3014-3024. Haverkamp, L.J., Anatomical and physiological development of the Xenopus embryonic motor system in the absence of neural activity, J. Neurosci., 6 (1986) 1338-1348. Hockfield, S., Kalb, R.G., Zaremba, S. and Fryer, H.J., Expression of neural proteoglycan correlates with the acquisition of mature neuronal properties in the mammalian brain, Cold Spring Harbor Symp. Quant. BioL, 55 (1990) 505-514. Hockfield, S., Scheetz, A.J. and Dubin, M.W., Monocular action potential blockade reduces expression of the Cat-301 proteoglycan in the LGN of neonatal, but not adult, cats, ARVO, 332 (1991) 1035. Hockfield, S. and Sur, M., Monoclonal antibody Cat-301 identifies Y-cells in the dorsal lateral geniculate nucleus of the cat, J. Camp. Neural., 300 (1990) 320-330. Jahr, C.E. and Yoshioka, K., Ia afferent excitation of motoneurons in the in vitro new-born rat spinal cord is selectively antagonized by kynurenate, J. Physiol., 370 (1986) 515-530. Jordan, C.L., Letinsky, M.S. and Arnold, A.P., The role of gonadal hormones in neuromuscular synapse elimination in rats. II. Multiple innervation persists in the adult levator ani muscle after juvenile androgen treatment, J. Neurosci., 9 (1989) 239-247. Jordan, C.L., Letinsky, MS. and Arnold, A.P., The role of gonadal hormones in neuromuscular synapse elimination in rats. I. Androgen delays the loss of multiple innervation in the levator ani muscle, J. Neurosci., 9 (1989) 229-238. Kalb, R.G. and Hockfield, S., Molecular evidence for early activity-dependent development of hamster motor neurons, J. Neurosci., 8 (1988) 2350-2360. Kalb, R.G. and Hockfield, S., Large diameter primary afferent input is required for expression of the Cat-301 proteoglycan on the surface of motor neurons, Neuroscience, 34 (1990) 391-401. Kalb, R.G. and Hockfield, S., Induction of a neuronal proteoglycan by the NMDA receptor in the developing spinal cord, Science, 250 (1990) 294-296. Kleinschmidt, A., Bear, M.F. and Singer, W., Blockade of ‘NMDA’ receptors disrupts experience-dependent plasticity of kitten striate cortex, Science, 238 (1987) 355-358. Lichtman, J.W., Magrassi, L. and Purves, D., Visualization of neuromuscular junctions over periods of several months in living mouse, J. Neurosci., 7 (1987) 1215-1222. Lnenicka, G.A. and Atwood, H.L., Age-dependent long-term adaptation of crayfish phasic motor axon synapses to altered activity, J. Neurosci., 5 (1985) 459-467. Lnenicka, G.A., Atwood, H.L. and Marin, L., Morphological transformation of synaptic terminals of a phasic motoneuron by long-term tonic stimulation, J. Neurosci., 6 (1986) 2252-2258. Lnenicka, G.A., Hong, S.J., Combatti, M. and LePage, S., Activity-dependent development of synaptic varicosities at crayfish motor terminals, J. Neurosci., 11 (1991) 1040-1048. Mendelson, B. and Frank, E., Specific monosynaptic sensory-motor connections form in the absence of patterned neural activity and motoneuronal cell death, .T.Neurosci., 11 (1991) 1390-1403. Miller, K.D., Chapman, B. and Stryker, M.P., Infusion of an
289 NMDA receptor antagonist blocks visually driven cortical activity in cat visual cortex, Sot. Neurosci. Abstr., 13 (1987) 357. 31 Nelson, P.G., Fields, R.D., Yu, C. and Neale, E.A., Mechanisms involved in activity-dependent synapse formation in mammalian central nervous system cell cultures, J. Neurobiol., 21 (1990) 138-156. 32 Purves, D. and Lichtman, J.W., Principles of Neural Development,
antigen on Y-cells in the cat lateral geniculate nucleus is regulated by visual experience, J. Neurosci., 8 (1988) 874-882. 40 Udin, S.B. and Keating, M.J., Plasticity in a central nervous system pathway in Xenopus: Anatomic changes in the isthmotectal projection after larval eye rotation, J. Comp. Neurol., 203 (1981) 575-594. 41 Walton, K. and Fulton, B.P., Ionic mechanisms underlying the firing properties of rat neonatal motoneurons studied in vitro,
Sinauer Associates, Sunderland, MA, 1985. 33 Reh, T.A. and Constantine-Paton, M., Eye specific segregation requires activity in 3 eyed Rana pipiens, J. Neurosci., 5 (1985) 1132-1143. 34 Ronnevi, L.-O. and Conradi, S., Ultrastructural evidence for spontaneous elimination of synaptic terminal on spinal motoneurons in the kitten, Brain Res., 80 (1974) 335-339. 35 Salt, T.E. and Hill, R.G., Neurotransmitter candidates of primary afferent somatosensoty fibers, Neuroscience, 10 (1983) 1083-1103. 36 Sham, C.J., Impulse activity and the patterning of connections during CNS development, Neuron, 5 (1990) 745-756. 37 Smith, D.E., The effect of deafferentation on the postnatal development of Clarke’s nucleus in the kitten - a Golgi study,
on morphology and physiology of cells in the cat’s lateral geniculate body, .I. Neurophysiol., 26 (1963) 978-993. 43 Wigston, D.J., Remodeling of neuromuscular junctions in adult mouse soleus, J. Neurosci., 9 (1989) 639-647. 44 Wigston, D.J., Repeated in vivo visualization of neuromuscular junctions in adult mouse lateral gastocnemius, J. Neurosci., 10 (1990) 1753-1761. 45 Zaremba, S., Guimaraes, A., Kalb, R.G. and Hockfield, S., Characterization of an activity-dependent, neuronal surface proteoglycan identified with monoclonal antibody Cat-301, Neuron, 2
Brain Res., 74 (1974) 119-130. 38 Stryker, M.P. and Harris, W.A., Binocular
(1989) 1207-1219. 46 Ziskind-Conhaim,
impulse blockade prevents the formation of ocular dominance columns in cat visual tortes, .I. Neurosci., 6 (1986) 2117-2133. 39 Sur, M., Frost, D. and Hockfield, S., Expression of a cell surface
Neuroscience, 19 (1986) 669-683. 42 Wiesel, T.N. and Hubel, D.H., Effects of visual deprivation
L., NMDA receptors mediate poly- and monosynaptic potentials in motoneurons of rat embryos, J. Neurosci., 10 (1990) 125-135.
Note added in proof (September 30, 1992)
A role for NMDA receptors in postnatal maturation of motor neurons is further supported by our recent demonstration that there is a high level of NMDA receptors transiently expressed in the ventral horn of the developing spinal cord @lb, R.G., Lidow, M.S., Halsted, M.J. and Hockfield, S., NMDA-receptors are transiently expressed in the developing spinal cord ventral horn, PNAS, 89 (1992) 8502-8506).
Brain Research Reviews, 17 (1992) 283-289 0 1992 Elsevier Science Publishers B.V. Ah rights reserved 0165-0173/92/$05.00
BRESR 95000
Short Review
Activity-dependent
development of spinal cord motor neurons
Robert G. Kalb a and Susan Hockfield b n Department of Neurology and b Section of Neurobiology, Yale University School of Medicine, New Haven, CT 04510 (USA) (Accepted 14 July 1992)
Key words: Motor neuron; Spinal cord; Activity; N-Methyl-D-aspartate
receptor; Cat-301; Proteoglycan; Critical period
CONTENTS 1. Introduction..
283
................................................................................
2. Molecular correlates of activity-dependent
development
3. Molecular and anatomical evidence for activity-dependent 4. Involvement of the N-methyl-D-aspartate
.................................................. development of motor neurons
..........................
subtype of glutamate receptor in motor neuron development .................
284
.
285
.
286
5. Patterns of afferent activity can influence physiological properties of motor neurons through activation of the NMDA receptor 6. Concludingremarks
287 287
............................................................................
7. Summary ....................................................................................
288
Acknowledgements
...................................................................................
288
.........................................................................................
288
References
Noteaddedinproof
..................................................................................
1. INTRODUCTION
During the development of the mammalian central nervous system (CNS), the mature set of synaptic connections is generated by a variety of mechanisms over an extended period of time. The choice of initial synaptic sites and partners is thought to be governed by intrinsic cellular properties including trophic and other chemical cues. In many systems, this initial set of synaptic connections is exuberant and is refined to its mature complement, at least in part, by neuronal activity32. Activity-dependent developmental processes have been demonstrated to have a profound effect not only on synaptic connectivity, but also on the morphological and electrophysiological properties of neurons4.36. Ac-
289
tivity-dependent developmental processes take place largely during a circumscribed period in early postnatal life known as the ‘critical’ or ‘sensitive’ period. Understanding the characteristics and mechanisms of these early plastic processes and the subsequent loss of plasticity may have important implications not only for development, but also for the recovery of function after damage to the adult nervous system. The visual system of vertebrates has provided some of the most compelling demonstrations of activity-dependent development. Visually evoked neuronal activity in the early postnatal period has profound effects on the anatomy and physiology of neurons in the visual system of mammals and amphibia. Visually evoked neuronal activity controls the segregation of geniculo-
CorresPondence to: R. Kalb,Department of Neurology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA. Fax: (1) (203) 7855694.
284 cortical input representing right and left eyes into ocular dominance columns in the primary visual cortex of cats36,“8 the segregation of retinal inputs in the lateral geniculate nucleus (LGNJ6, the retinotopic mapping of afferents from the nucleus isthmi onto the tectum in two-eyed frogs4” and the segregation of retinal afferents into ocular dominance stripes in the tectum of three-eyed frogs3’. In all these systems, competition between two inputs with non-synchronous patterns of activity leads to the stabilization of a subset of connections4936. The cellular and molecular biological events underlying the selection and stabilization of mature synaptic organization are just beginning to be explored. Study of developing motor neurons has yielded conflicting results concerning the importance of activity in the acquisition of mature neuronal properties. Despite the profound changes in motor neuron structure that occur during the postnatal period3,34 few studies have described an activity-dependent critical period in motor neuron development. The stereotypical development of many motor neuron characteristics, such as the formation of initial connections between homonomous muscle afferents and their motor neurons’” and the formation of pattern generators for fictive swimming in is thought to be mediated by intrinsic amphibiansI signals and chemical cues. In the chicken, blockade of muscle activity by d-tubocurare application does not perturb the accurate mapping of primary afferents onto motor neurons, although the Ia afferent-induced motor neuron EPSP’s are twice normal sizez9. While these observations suggest that the initial matching of la afferents onto motor neurons is likely to be activityindependent, it does not preclude the existence of activity-dependent sculpting of the synaptic input to motor neurons during the postnatal period. Other characteristics of motor neurons clearly are not hard wired and are likely to be under the control of epigenetic factors. The hormonal control of synapse elimination from sexually dimorphic muscles is an example of a developmental phenomenon controlled by external influences during a critical period in early postnatal life 19-20 . Synapse elimination from the levator ani (LA) muscle in male rats occurs during the first month of postnatal life and is sensitive to manipulation of gonadal androgen levels. Removal of endogenous androgens in juvenile males by castration accelerates the loss of polyinnervation2’. Precocious exposure to elevated levels of androgens prevents the loss of polyinnervation and this polyinnervation persists for months after the cessation of exogenous androgen administration”. These results suggest that physiological levels of androgens during a critical period in early life
normally regulates the maturation of the LA neuromuscular system. Neuronal activity also has been demonstrated to influence the development of other motor neuron characteristics26-28. Synapses at the crustacean neuromuscular junction can be physiologically categorized as tonic or phasic based on neurotransmitter release characteristics and fatigue resistance. Tonic-patterned stimulation of single phasic motor neurons can transform the physiology of motor neurons from phase into more tonic-like. Such phasic-to-tonic transformation is accompanied by the acquisition of tonic morphology (larger terminal arbors with more varicosities) in phasic motor neurons 28. This activity-induced motor neuron plasticity is age-dependent, as its induction is significantly greater in young crayfish than in adults2’. These observations show that the pattern of activity of motor neurons in early life can lead to alterations in motor neuron properties. In the mammalian neuromuscular system, it has recently been demonstrated that patterns of activity (phasic versus tonic) can similarly influence the morphology of mouse DRG neurons in vitro’. Finally, although activity clearly has the potential to influence motor neuron characteristics, it is possible that not all motor neurons are equally susceptible to the effects of activity. In vivo imaging techniques have permitted repeated observations of the mouse neuromuscular junction over extended periods of time2s,43*44. Such studies suggest significant synaptic remodeling of the neuromuscular junction in some but not all adult muscles. The remodelling observed may be activity-dependent, since remodelling was more pronounced in slow-twitch (tonic) muscle fibers than in fast-twitch (phasic) fibers. These differences in the plasticity of muscle types in adults may reflect lifelong differences between motor neurons and their targets. Some types of motor neurons may be embryonically hard-wired while others may be more sensitive to activity-dependent structural and functional modifications. 2. MOLECULAR CORRELATES OF ACTIVI-N-DEPENDENT DEVELOPMENT Our work over the last several years has been based on the hypothesis that changes in, or the acquisition of new, molecular properties might accompany the acquisition of mature synaptic structure. In the cat visual system we have demonstrated that the chondroitin sulfate proteoglycan recognized by monoclonal antibody Cat-3014s is expressed late in the postnatal period and is regulated by neuronal activity during early postnatal life39. In the cat LGN, the cell-surface Cat-301 proteoglycan is expressed on Y-cells in animals raised
285 in a normal visual environment “, appearing between postnatal days 30 and 90 (ref. 39). In animals deprived of normal visual experience by dark rearing13 monocular lid suture39, or intraocular administration of tetrodotoxin’6, the expression of the Cat-301 proteoglycan is dramatically reduced. Visual deprivation in adult animals for extended periods of time has no effect on expression of the proteoglycan. The Cat-301 antigen is, therefore, a unique positive molecular marker of normal activity-dependent development in the cat visual system. 3. MOLECULAR AND ANATOMICAL EVIDENCE FOR ACTIVITY-DEPENDENT
DEVELOPMENT OF MOTOR
NEURONS
We have used Cat-301 to explore the possibility that activity-dependent development might also be a feature of spinal motor neurons. Cat-301 immunohistochemistry combined with retrograde labelling from the sciatic nerve demonstrated that in normal animals all sciatic motor neurons express the Cat-301 antigen*l. In the spinal cord, like in the LGN, Cat-301 immunoreactivity develops relatively late in the postnatal period. Immunoreactivity is not present at birth and
adult levels of Cat-301 expression are not reached until P14 (that is, all sciatic motor neurons are Cat-301 positive at P14). Two other motor neuron antigens (recognized by monoclonal antibodies Rat-302 and Rat-303) are also first expressed during the postnatal period; they, too, are not present at birth and develop to adult levels during the second postnatal week of life*l. Regulation of Cat-301 expression on motor neurons mimics the activity-dependent expression demonstrated in the cat LGN. Connectivity and activity of motor neurons was disrupted in our initial experiments by a quite severe lesion, sciatic nerve crush*l (see Fig. 1). The effect of nerve crush performed at P5 (several days before normal Cat-301 expression) was compared to the effect of nerve crush in adult animals. In animals with sciatic nerve crush at P5, Cat-301 expression was profoundly inhibited: less than 5% of sciatic motor neurons expressed the Cat-301 antigen at P21 (Fig. 1). Importantly, this reduction in Cat-301 expression was relatively specific in that the antigens recognized by Rat-302 and Rat-303 showed no reduction in expression ‘l. Sciatic nerve lesions in adult animals had no effect on Cat-301 expression, demonstrating that antigen expression is not mediated simply by continuous
Fig. 1. Two weeks after neonatal sciatic nerve lesion at P7, retrogradely labeled motor neurons do not express Cat-301 immunoreactivity. A: one motor neuron in this field shows retrograde labeling with Fast blue (arrowhead), 2 weeks after a sciatic nerve lesion. The UV illumination faintly excites the FITC labeling of a Cat-301 positive, Fast blue negative cell in the lower portion of the panel (arrow). Under UV illumination this cell fluoresces a faint yellow-green. B: same field shown in (A) viewed under FITC optics. No surface-associated Cat-301 immunoreactivity (visualized with FITC-conjugated secondary antibody) is seen at the location of the Fast blue cell (arrowhead). As this photomicrograph has been overexposed to more clearly demonstrate the lack of Cat-301 immunoreactivity, the fine intracellular particulate Fast blue staining is seen breaking though under FITC illumination. A Fast blue negative, Cat-301 positive cell is seen in the lower portion of the field (arrow). x390. Taken from Ref. 21.
286 4. INVOLVEMENT
neuronal activity, but specifically by activity during a critical period in development. Subsequent experiments demonstrated that the effect of sciatic nerve crush on Cat-301 expression was not due to direct injury to the motor neuron itself, but was more likely due to a disruption in the pattern of activity of the motor neuron. Lesions of descending input to motor neurons 2’ (by thoracic hemicordotomy) or of primary afferent input22 (by dorsal rhizotomy) also produced reductions in Cat-301 immunoreactivity when performed on early postnatal animals. Thus, lesions that altered input to motor neurons, without directly injuring the motor neurons, decreased Cat-301 expression (Fig. 2). These results suggest that the coordinated input from upper motor neurons and segmental afferents during early postnatal life is required for the normal molecular development of motor neurons. These results parallel observations in the visual system where patterns of neuronal activity in early postnatal life determine the course of neuronal maturation. One set of experiments was designed to determine whether a particular class of primary afferent input was required for Cat-301 expression. Capsaicin, a neurotoxin selective for small diameter afferents, administered to neonates had no effect on Cat-301 expression22, providing evidence that expression of the Cat301 proteoglycan on motor neurons depends upon large diameter primary afferents. Many of these fibers are glutamatergic35.
Thoracic
One class of glutamate receptor, the NMDA receptor, has been linked to developmentally regulated synaptic plasticity in the visual system of cat’,24 and frog 2,4. We have tested whether the NMDA receptor might play a role in the development of motor neurons. Administration of the NMDA antagonists APV or MK-801 during the neonatal period inhibited the expression of the Cat-301 proteoglycan on motor neurons23 (summarized in Fig. 2). This inhibition was dose-dependent, stereospecific and selective for the Cat-301 proteoglycan. The inhibition was also temporally restricted, in that in adult animals Cat-301 immunoreactivity on motor neurons was not diminished by NMDA receptor blockade. The effect of NMDA receptor blockade on motor neuron development is most likely to be mediated by receptors at the segmental level because the efficacy of antagonists decreased in rough proportion to the distance from the sciatic motor neuron pool at which they were administered2’. These results suggest that motor neurons undergo an activity-dependent, NMDA-mediated critical period in development. The role of NMDA receptor in development of the mammalian visual system currently is a topic of debate. In the developing ferret LGN (M. Sur, personal com-
Hemlcordotomy
Capsalcln
Sciatic antagonlsm
N-METHYL-D-ASPAR-
TOR NEURON DEVELOPMENT
Dorsal Rhlzotomy
NMDA receptor
OF THE
TATE SUBTYPE OF GLUTAMATE RECEPTOR IN MO-
- 49% - 45%
- 98%
Nerve Crush - 4%
- 22%
Fig. 2. Influence of neonatal lesions on motor neuron Cat-301 expression. Surgical and pharmacologic lesions of different components of the neuromuscular system was performed on postnatal day 7 (P7) and at P21 Cat-301 immunoreactivity on sciatic motor neurons was assessed. The percentages reported represent the average number of Cat-301 positive motor neurons after each manipulation. Each manipulation was performed on at least three separate animals from different litters and the Cat-301 status was derived by assaying a minimum of 300 retrogradely labeled sciatic motor neurons. Of note, in parallel with the neonatal lesion experiments, adult animals were subjected to the same interventions and the Cat-301 status of sciatic motor neurons was assayed 2 weeks later. When any of the above lesions were performed on adult animals, in all instances, virtually all motor neurons retained Cat-301 immunoreactivity.
287 munication) and cat visual cortex3’ a large component (and perhaps the majority) of the excitatory drive to neurons is mediated by activation of the NMDA receptor, so that NMDA receptor blockade may be equivalent to a non-specific inhibition of neuronal activity’. In contrast to the visual system, experiments on spinal cord neurons suggest that the major excitatory drive to motorneurons during development is not mediated by activation of the NMDA receptor and that a substantial amount of excitation to motorneurons is carried by non-NMDA glutamate receptors in the pre- and early . We have demonstrated that post-natal period 18*46 spinal cord motor neurons can be driven by segmental reflex pathways in the presence of NMDA receptor antagonists I5. NMDA antagonists do not, therefore, globally prevent synaptic activation of motor neurons. Together these results suggest that coordinated action of primary afferents and descending inputs can activate the NMDA receptor at the segmental level (see Note added in proof). Activation of NMDA receptors on target cells may then regulate the expression of a class of neuronal proteins, of which Cat-301 is one example, that might subserve the anatomical and physiological consequences of activity-dependent development. 5. PATTERNS OF AFFERENT ACTIVITY CAN INFLUENCE PHYSIOLOGIC PROPERTIES OF MOTOR NEURONS THROUGH ACTIVATION
OF THE NMDA RE-
CEPTOR
Our in vivo studies provide molecular evidence that normal motor neuron development requires the functioning of NMDA receptors at the spinal segmental level. These observations complement in vitro observations of NMDA receptor mediated changes in synaptic efficacy of ventral horn neurons8Y31. Dissociated cells from the spinal ventral horn (VII) were cultured in the central chamber of a three compartment tissue culture dish while dorsal root ganglion (DRG) neurons were cultured in the two flanking chambers and allowed to grow axons into the central chamber. One of the DRG chambers was chronically stimulated at a level sufficient to drive the VII neurons. This led to an enhanced synaptic efficacy and a larger number of inputs from the stimulated cells in comparison to the inputs from unstimulated control side. The increase in efficacy could be prevented by application of NMDA receptor antagonists during the stimulation period. Further work has shown that this process is likely to be mediated by a calcium conductance’. These experiments demonstrate that the DRG-VH synapses possess the potential for activity-dependent modification. They suggest that
the developing spinal cord, like the developing visual system, may utilize activity-dependent mechanisms to establishing synaptic connectivity.
6. CONCLUDING REMARKS
Studies of neuromuscular system development in many species have shown that extensive electrophysiologica1’2T41and morphological5 changes occur during the early postnatal period. The studies described here provide support for the possibility that the maturation of motor neurons, like neurons in many other systems, has an activity-dependent component. Whether other motor neuron characteristics might be regulated by neuronal activity during development remains to be explored in detail. In the LGN”*‘3,42 and the spinal nucleus of Clarke37, cell soma size and dendritic morphology are altered by activity-deprivation during a critical period in early development. In our ongoing experiments, measurements of retrogradely labeled motor neurons in normal and in animals deprived of normal neuromuscular activity during the neonatal period indicate that activity-deprived motor neurons are significantly smaller in cross-sectional area than control motor neurons. Since cell size and dendritic architecture frequently co-vary, it will be interesting to see if the dendritic arbors of motor neurons are also regulated by neuronal activity during a critical period in early postnatal life. Why motor neurons might have an activity-dependent period in development remains open to speculation at present. Based on insights gleaned from the visual system, one possibility is that activity-dependent processes are required to fine tune the inputs (or their gain) onto motor neurons for optimal neuromuscular performance. A child’s walking, running and jumping might provide a pattern of activation of primary afferent and descending inputs onto motor neurons (and/or a set of interneurons) that could lead to activation of NMDA receptors and activity-dependent modifications of the motor neuron. One might then expect that the activity-driven shaping of the inputs to and response properties of motor neurons would be reflected in dendritic architecture as well as synaptology. The activity-dependent fine tuning of the neuromuscular system might account, in part, individual differences in motor performance. One implication of the existence of a critical period in motor neuron maturation is that the structural and functional adaptability of motor neurons to changing demands has finite temporal limits, which may bear on the functional outcome after injury to the motor system.
288 7. SUMMARY Patterned neuronal activity in early postnatal life can regulate the acquisition of the mature morphological and electrophysiological properties of neurons. Many properties of motor neurons are developmentally regulated and may be influenced by epigenetic factors. The pattern of activation of motor neurons can regulate axon terminal morphology and synaptic efficacy at the neuromuscular junction. Motor neuron morphology and synaptic connections can also be modified by exposure to specific hormones in the early postnatal period. The acquisition of mature physiological and anatomical properties is paralleled by the acquisition of specific molecular properties. Recent experiments using molecular markers for motor neuron differentiation indicate that motor neurons undergo activity-dependent development during a circumscribed period in early postnatal life. Normal motor neuron differentiation requires a normal pattern of neuronal activity in early postnatal life. Differentiation also requires activation of the NMDA receptor over the same time period. The activity-dependent development of morphological, electrophysiological and molecular properties of motor neurons is similar to activity-dependent development in the vertebrate visual system. The neuromuscular system may provide an accessible system for characterizing the molecules subserving the translation of patterned neuronal activity into mature neuronal phenotype. Acknowledgements. This work was supported by Grants to R.G.K. from NIH (NS012471, the Muscular Dystrophy Association and the Daniel Heumann Fund for Spinal Cord Research and to S.H. from NIH (EY-06511) and NSF (BNS-8544681).
REFERENCES 1 Bear, M.F., Kleinschmidt, A., Gu, Q. and Singer, W., Disruption of experience-dependent synaptic modifications in striate cortex by infusion of an NMDA receptor antagonist, J. Neurosci., 10 (1990) 909-925. 2 Cline, H.T., Debski, E.A. and Constantine-Paton, M., N-methylD-aspartate receptor antagonist desegregates eye-specific columns, PNAS, 84 (1987) 4342-4345. 3 Conradi, S. and Ronnevi, L.-O., Ultrastructure and synaptology of initial axon segment of cat spinal motor neurons during early postnatal development, J. Neurocytol., 6 (1977) 195-210. 4 Constantine-Paton, M., Cline, H.T. and Debski, E., Patterned activity, synaptic convergence and the NMDA receptor in developing visual pathways, Annu. Reu. Neurosci., 13 (1990) 129-154. 5 Cummings, J.P. and Stelzner, D.J., Prenatal and postnatal development of lamina IX neurons in the rat thoracic spinal cord, fin. Neural., 83 (1984) 155-166. 6 Dubin, M.W., Stark, L.A. and Archer, S.M., A role for actionpotential activity in the development of neuronal connections in the kitten retinogeniculate pathway, J. Neurosci., 6 (1986) 102l1036. 7 Fields, R.D., Neale, E.A. and Nelson, P.G., Effects of patterned electrical activity on neurite outgrowth from mouse sensory neurons, J. Neurosci., 10 (1990) 2950-2964.
8 Fields, R.D., Yu, C. and Nelson, P.G., Calcium, network activity 9 10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
and the role of NMDA channels in synaptic plasticity in vitro, .I. Neurosci., 11 (1991) 134-146. Frank, E., The influence of neuronal activity on patterns of synaptic connections, TINS, 10 (1987) 188-190. Frank, E. and Jackson, P.C., Normal electrical activity is not required for the formation of specific sensory-motor synapses, Bruin Res., 378 (1986) 147-151. Friedlander, M.J., Stanford, L.R. and Sherman, S.M., Effects of monocular deprivation on the structure-function relationship of individual neurons in the cat’s lateral geniculate nucleus, J. Neurosci., 2 (1982) 321-330. Fulton, B.P. and Walton, K., Electrophysiological properties of neonatal rat motoneurons studied in vitro, J. PhysioL, 370 (1986) 651-678. Guimaraes, A., Zaremba, S. and Hockfield, S., Molecular and morphological changes in cat lateral geniculate nucleus and visual cortex induced by visual deprivation are revealed by monoclonal antibodies Cat-301 and Cat-304, J. Neurosci., 10 (1990) 3014-3024. Haverkamp, L.J., Anatomical and physiological development of the Xenopus embryonic motor system in the absence of neural activity, J. Neurosci., 6 (1986) 1338-1348. Hockfield, S., Kalb, R.G., Zaremba, S. and Fryer, H.J., Expression of neural proteoglycan correlates with the acquisition of mature neuronal properties in the mammalian brain, Cold Spring Harbor Symp. Quant. BioL, 55 (1990) 505-514. Hockfield, S., Scheetz, A.J. and Dubin, M.W., Monocular action potential blockade reduces expression of the Cat-301 proteoglycan in the LGN of neonatal, but not adult, cats, ARVO, 332 (1991) 1035. Hockfield, S. and Sur, M., Monoclonal antibody Cat-301 identifies Y-cells in the dorsal lateral geniculate nucleus of the cat, J. Camp. Neural., 300 (1990) 320-330. Jahr, C.E. and Yoshioka, K., Ia afferent excitation of motoneurons in the in vitro new-born rat spinal cord is selectively antagonized by kynurenate, J. Physiol., 370 (1986) 515-530. Jordan, C.L., Letinsky, M.S. and Arnold, A.P., The role of gonadal hormones in neuromuscular synapse elimination in rats. II. Multiple innervation persists in the adult levator ani muscle after juvenile androgen treatment, J. Neurosci., 9 (1989) 239-247. Jordan, C.L., Letinsky, MS. and Arnold, A.P., The role of gonadal hormones in neuromuscular synapse elimination in rats. I. Androgen delays the loss of multiple innervation in the levator ani muscle, J. Neurosci., 9 (1989) 229-238. Kalb, R.G. and Hockfield, S., Molecular evidence for early activity-dependent development of hamster motor neurons, J. Neurosci., 8 (1988) 2350-2360. Kalb, R.G. and Hockfield, S., Large diameter primary afferent input is required for expression of the Cat-301 proteoglycan on the surface of motor neurons, Neuroscience, 34 (1990) 391-401. Kalb, R.G. and Hockfield, S., Induction of a neuronal proteoglycan by the NMDA receptor in the developing spinal cord, Science, 250 (1990) 294-296. Kleinschmidt, A., Bear, M.F. and Singer, W., Blockade of ‘NMDA’ receptors disrupts experience-dependent plasticity of kitten striate cortex, Science, 238 (1987) 355-358. Lichtman, J.W., Magrassi, L. and Purves, D., Visualization of neuromuscular junctions over periods of several months in living mouse, J. Neurosci., 7 (1987) 1215-1222. Lnenicka, G.A. and Atwood, H.L., Age-dependent long-term adaptation of crayfish phasic motor axon synapses to altered activity, J. Neurosci., 5 (1985) 459-467. Lnenicka, G.A., Atwood, H.L. and Marin, L., Morphological transformation of synaptic terminals of a phasic motoneuron by long-term tonic stimulation, J. Neurosci., 6 (1986) 2252-2258. Lnenicka, G.A., Hong, S.J., Combatti, M. and LePage, S., Activity-dependent development of synaptic varicosities at crayfish motor terminals, J. Neurosci., 11 (1991) 1040-1048. Mendelson, B. and Frank, E., Specific monosynaptic sensory-motor connections form in the absence of patterned neural activity and motoneuronal cell death, .T.Neurosci., 11 (1991) 1390-1403. Miller, K.D., Chapman, B. and Stryker, M.P., Infusion of an
289 NMDA receptor antagonist blocks visually driven cortical activity in cat visual cortex, Sot. Neurosci. Abstr., 13 (1987) 357. 31 Nelson, P.G., Fields, R.D., Yu, C. and Neale, E.A., Mechanisms involved in activity-dependent synapse formation in mammalian central nervous system cell cultures, J. Neurobiol., 21 (1990) 138-156. 32 Purves, D. and Lichtman, J.W., Principles of Neural Development,
antigen on Y-cells in the cat lateral geniculate nucleus is regulated by visual experience, J. Neurosci., 8 (1988) 874-882. 40 Udin, S.B. and Keating, M.J., Plasticity in a central nervous system pathway in Xenopus: Anatomic changes in the isthmotectal projection after larval eye rotation, J. Comp. Neurol., 203 (1981) 575-594. 41 Walton, K. and Fulton, B.P., Ionic mechanisms underlying the firing properties of rat neonatal motoneurons studied in vitro,
Sinauer Associates, Sunderland, MA, 1985. 33 Reh, T.A. and Constantine-Paton, M., Eye specific segregation requires activity in 3 eyed Rana pipiens, J. Neurosci., 5 (1985) 1132-1143. 34 Ronnevi, L.-O. and Conradi, S., Ultrastructural evidence for spontaneous elimination of synaptic terminal on spinal motoneurons in the kitten, Brain Res., 80 (1974) 335-339. 35 Salt, T.E. and Hill, R.G., Neurotransmitter candidates of primary afferent somatosensoty fibers, Neuroscience, 10 (1983) 1083-1103. 36 Sham, C.J., Impulse activity and the patterning of connections during CNS development, Neuron, 5 (1990) 745-756. 37 Smith, D.E., The effect of deafferentation on the postnatal development of Clarke’s nucleus in the kitten - a Golgi study,
on morphology and physiology of cells in the cat’s lateral geniculate body, .I. Neurophysiol., 26 (1963) 978-993. 43 Wigston, D.J., Remodeling of neuromuscular junctions in adult mouse soleus, J. Neurosci., 9 (1989) 639-647. 44 Wigston, D.J., Repeated in vivo visualization of neuromuscular junctions in adult mouse lateral gastocnemius, J. Neurosci., 10 (1990) 1753-1761. 45 Zaremba, S., Guimaraes, A., Kalb, R.G. and Hockfield, S., Characterization of an activity-dependent, neuronal surface proteoglycan identified with monoclonal antibody Cat-301, Neuron, 2
Brain Res., 74 (1974) 119-130. 38 Stryker, M.P. and Harris, W.A., Binocular
(1989) 1207-1219. 46 Ziskind-Conhaim,
impulse blockade prevents the formation of ocular dominance columns in cat visual tortes, .I. Neurosci., 6 (1986) 2117-2133. 39 Sur, M., Frost, D. and Hockfield, S., Expression of a cell surface
Neuroscience, 19 (1986) 669-683. 42 Wiesel, T.N. and Hubel, D.H., Effects of visual deprivation
L., NMDA receptors mediate poly- and monosynaptic potentials in motoneurons of rat embryos, J. Neurosci., 10 (1990) 125-135.
Note added in proof (September 30, 1992)
A role for NMDA receptors in postnatal maturation of motor neurons is further supported by our recent demonstration that there is a high level of NMDA receptors transiently expressed in the ventral horn of the developing spinal cord @lb, R.G., Lidow, M.S., Halsted, M.J. and Hockfield, S., NMDA-receptors are transiently expressed in the developing spinal cord ventral horn, PNAS, 89 (1992) 8502-8506).
