- Email: [email protected]
An attempt to localize the lumbar locomotor generator in the rabbit using 2-deoxy-[14C]glucose autoradiography
Neuroscience Letters, 86 (1988) 139-143 Elsevier Scientific Publishers Ireland Ltd.
139
NSL 05196
An attempt to localize the lumbar locomotor generator in the rabbit using 2-deoxy-[14C]glucose autoradiography D. Viala 1, C. Buisseret-Delmas 2 a n d J.J. Portal 1 ILaboratoire de Neurophysiologie, UA C.N.R.S. 1199, Facult~ des Sciences Mirande, Dijon (France) and 2Unitb de Recherche Neuropharmacologique de I'INSERM, Paris (France) (Received 20 October 1987; Accepted 2 November 1987)
Key words." Lumbar locomotor generator; 2-Deoxyglucose; Rabbit An attempt was made to find the anatomical localization of the lumbar locomotion generators using 2-deoxy-[14C]glucose (2-DG) uptake in acute low spinal preparation of rabbits unanaesthetized, curarized and injected with nialamide and dihydroxyphenylalanine (DOPA). In such conditions, the locomotor generators were forced to work in isolation for 45 min without interruption as attested by the rhythmic activity recorded in hindlimb muscle nerves. Comlbared to spinal control preparations not activated pharmacologically, the treated animals showed a specific labeling in the intermediate part of the grey matter, extending from L6 to St.
It is well documented that the lumbar cord, deprived of supraspinal and peripheral informations, can generate locomotor-like activities, due to the presence of intrinsic generators [2, 7, 12]. During effective or fictive locomotor activity, lumbar neurones display bursty firing related to one phase of the ipsilateral locomotor cycle. These neurones are found in the intermediate and in the ventral part of the lumbar grey matter [3-5, 10]. Taking into account the dorsoventral sensorimotor organization of the spinal cord, one can assume that the more dorsally located bursty firing neurones are part of the lumbar locomotion generators. On the other hand, the rostrocaudal extent of such left and right generators is, up to date, not known. The present work aims to localize anatomically these generators using 2-deoxy-[14C]glucose (2-DG) uptake [11], a technique that allowed Batini et al. [1] to label the olivary generator responsible for harmaline tremor. We have used a preparation in which the lumbar generators, isolated from supraspinal levels and peripheral feedback are forced into continuous bursty firing by treatment with nialamide-dihydroxyphenylalanine (DOPA) [121. Seven young rabbits (500-1000 g body wt.) were used for the experiments. After Correspondence." D. Viala, Laboratoire de Neurophysiologie, UA C.N.R.S. 1199, Facult6 des Sciences Mirande, F-21004 Dijon, France. 0304-3940/88/$ 03.50 O 1988 Elsevier Scientific Publishers Ireland Ltd.
140
general anaesthesia with sodium pentobarbital (26 mg/kg i.v.) and tracheotomy, a catheter was inserted into one jugular vein for further injections. An extensive decortication was performed after craniectomy. Nerves to tibialis anterior (TA) and gastrocnemius medialis (GM) muscles were isolated and mounted on a pair of silver electrodes for recording of their efferent activities. A laminectomy was performed at T~2 level and the spinal cord was transected after local anaesthesia by lidocaine infiltration: the lumbar cord was thus isolated from supraspinal influences. The animals were then immobilized with flaxedil (to suppress peripheral feedback) and artificially ventilated. In 5 animals, nialamide (100 mg/kg i.v.) was injected just after general anaesthesia. M o t o r activities were recorded in hindlimb muscle nerves after administration of D O P A (100 mg/kg i.v.), 6 h after the onset of the experiment (this delay being necessary for the preparation to recover from general anaesthesia and for nialamide to become efficient). The efferent nerve activities were continuously monitored from 15 to 60 min (t~5 t60) after D O P A injection and stored in a tape recorder. The same protocol applies to the two other animals not injected with nialamide-DOPA and used as controls. All the animals received 100/~Ci/kg of [14C]2-DG i.v. at tls. They were sacrificed at t60 with a nembutal overdose. After extensive laminectomy, the lumbar cord was rapidly removed. Landmarks of vertebral limits were previously performed in order to determine thereafter the position of the spinal segments. The spinal segments L6 to $1 were taken off and frozen in isopentane precooled from - 4 5 to - 5 0 ° C . Serial sagittal or cross-sections (16/~m thick) were performed in a cryostat and were then rapidly dried at 40°C. The sections were exposed for autoradiography on X-ray film (single-coated Kodirex or single-coated Agfa Mamoray) for a few days. Thereafter the sections were stained for Nissl substance. In the control animals in which the lumbar locomotion generator was deprived of supraspinal and segmental control, the corresponding motoneurones were silent and no efferent activity could be disclosed in the hindlimb muscle nerves (Fig. I A). On the contrary, in the 5 experimental rabbits treated with nialamide-DOPA, continuous locomotor-like burst activity at the frequency of about 1 Hz [12] was observed in the muscle nerves throughout the 45 min of recording from tl5 to/60 (Fig. 1A'). The distribution of 2 - D G radioactivity was traced in the lumbosacral cords of the treated and of the control animals. The differences observed are shown in the crosssections of Fig. !B. In the control (a'), 3 zones of discrete labeling are met, due to different neurone density, corresponding to the substantia gelatinosa in the dorsal horn, to the central grey and to the motoneurone pools in the ventral horn (arrows in a and a'). The same regions of the cord are in general more labeled in treated animals. Moreover, there are two additional zones of specific labeling: one corresponding to the dorsal pool of motoneurones (horizontal arrow in b') and the second one corresponding to the intermediate part of the grey matter (coupled arrows in b'). Latter zone appears as a strip which extends from the dorsal pool of motoneurones (Mns) towards the medial border of the grey matter. Labeling of the dorsal pool of motoneurones is much more important than that of the ventral pool although both
141
Fig. 1. Labeling of the L7-S1 spinal cord in acute spinal preparations. A and A': recording of the spontaneous efferent activity in the tibialis anterior (TA) and gastrocnemius medialis (GM) muscle nerves in a control spinal preparation (A) and in another spinal preparation after nialamide-DOPA administration (A', 20 min after DOPA injection). In B, a,a' and b,b' show the labeled cross-sections at SI level (a and b) and their radioautography (a' and b') in a control preparation (a and a') and in a preparation injected with DOPA (b and b'). C: parasagittal slice of a spinal-DOPA preparation. Stained section (c) and its radioautography (c') are presented below each other. The vertical black line delimitates the substantia gelatinosa (its size is due to a torsion of this part of the cord) and the arrow indicates the maximal labeling in the intermediate grey matter.
142
dorsal and ventral Mns innervate hindlimb muscles and have been shown to be active during fictive spinal locomotion [8]. When observed with accuracy, the labeling in the region of the dorsal pool is not limited to the Mns but extends more medially. Therefore the labeling of this region is presumably due not only to the few motoneurones but also to other neighbouring neurones of the intermediate part of the cord. We have attempted to confirm this assumption and to determine the caudorostral extension of the labeling using sagittal serial sections performed at levels L 6 - S 1. In control preparations the findings were the following: in the lateral sections, light grey matter labeling corresponds to the substantia gelatinosa dorsally, and to the dorsal and ventral pools of motoneurones ventrally, these two structures being separated by almost unlabeled white matter; in the medial sections, light labeling is uniform and Mn label is undistinguishable from the surrounding. In the pharmacologically treated preparations, a general more important labeling of the grey matter is met. In addition, a specifically labeled zone is observed ventral to the substantia gelatinosa (latter delimited by a black vertical line on Fig. l c and c') and separated from it by an unlabeled band of 0.5 mm. This zone extends longitudinally from the caudal part of Sj to the caudal part of L 6 and it can be followed for 1.0 mm in the lateromedial direction. It clearly corresponds to the zone of intense labeling observed in the cross sections. The maximal density is at caudal level where the strip ends enlarged at a place where there is no histological substrate of increased neuronal density (wide arrow in Fig. 1Ccc'). These results show that the specific labeling does not correspond to the dorsal pool of Mns (the dorsal pool, which provides ankle and toe muscles, and the ventral pool that provides more proximal hindlimb muscles being equally labeled) but to an intermediate zone whose lateral edge is close to the dorsal pool of Mns. Because the maximal intensity of labeling corresponds to the intermediate grey matter where the more dorsal bursty firing neurones have been recorded during electrophysiological investigations (unpublished data), we may conclude that the neurones in this zone are active during locomotion and therefore belong to the central pattern generator, and that these active neurones have a higher metabolism than the follower neurones (including the Mns) during locomotion. The neurones which belong to the central pattern generator and the motoneurones are both active during locomotion. The reason why the generator neurones have a higher metabolism than the motoneurones cannot be answered with the present experiments: perhaps their density is higher or their firing level is much more increased than the motoneuronal one. Else this may possibly sign special metabolic requirements of these neurones if they are endogenous oscillators or pacemakers: this assumption cannot be excluded since some isolated neurones from the Mammalian spinal cord, whose function is still unknown, have been identified as endogenous oscillators [9]. Griilner [6] proposed an organization of the lumbar locomotion generator consisting in independent subgenerators coupled by interconnections. Our results are not in contradiction with this since the specifically labeled regions are not uniformly distributed within the caudo-rostral part of the cord. In addition, they suggest that the caudal part of the presumed generator (arrow in Fig. lCc') may play ~ special role of leader of the subgenerators.
143 We are grateful to Cesira Batini for her valuable comments
and suggestions on
this manuscript. 1 Batini, C., Buisseret-Delmas, C. and Conrath-Verrier, M., Harmaline induced tremor. I-Regional metabolic activity as revealed by ~4C-2-deoxyglucose in cat, Exp. Brain Res., 42 (1981) 371-382. 2 Brown, T.G., The intrinsic factors in the act of progression in the Mammal, Proc. R. Soc. Lond. B, 84 (1911) 308 319. 3 Edgerton, V.R., Grillner, S. and Sjostrom, A., On the spinal stepping generator, Soc. Neurosci. Abstr., I (1975) 615. 4 Edgerton, V.R., Grillner, S., Sjostrom, A. and Zangger, P., Central generation of locomotion in Vertebrates. In R.M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural Control of Locomotion, Vol. 18, New York, 1976, pp. 439~,64.. 5 Feldman, A.G. and Orlovsky, G.N., Activity of interneurones mediating reciprocal inhibition during locomotion, Brain Res., 84 (1975) 181 194. 6 Grillner, S., Control of locomotion in bipeds, tetrapods and fish. In J.M. Brookhart, V.B. Mountcastle, V.B. Brooks and S.R. Geiger (Eds.), Handbook of Physiology, Section 1: The Nervous System, Vol. II, Part 2, American Physiological Society, Bethesda, 1981, pp. 1179-1236. 7 Grillner, S. and Zangger, P., Locomotor movements generated by the deafferented spinal cord, Acta Physiol. Scand., 91 (1974) 38-39A. 8 Grillner, S. and Zangger, P., On the central generation of locomotion in the low spinal cat, Exp. Brain Res., 34 (1979) 241-261. 9 Legendre, P., McKenzie, J.S., Dupouy, B. and Vincent, J.D., Evidence for bursting pacemaker neurones in cultured spinal cord cells, Neuroscience, 16 (1985) 753 767. 10 Orlovsky, G.N. and Feldman, A.G., Classification of lumbo-sacral neurons according to their discharge patterns during evoked locomotion, Neurophysiology, 4 (1972) 410-417. 11 Sokoloff, L., Reivich, M., Kennedy, C., Des Rosiers, M.H., Patlak, C.S., Pettigrew, K.D., Sakurade, O. and Shinohara, M., The ~4C-deoxyglucose method for measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat, J. Neurochem., 28 (1977) 897 916. 12 Viala, D. and Buser, P., Modalit6s d'obtention de rythmes locomoteurs chez le lapin spinal par traitements pharmacologiques (DOPA, 5-HTP, D-amph&amine), Brain Res., 35 (I 971) 15 l - 165.
139
NSL 05196
An attempt to localize the lumbar locomotor generator in the rabbit using 2-deoxy-[14C]glucose autoradiography D. Viala 1, C. Buisseret-Delmas 2 a n d J.J. Portal 1 ILaboratoire de Neurophysiologie, UA C.N.R.S. 1199, Facult~ des Sciences Mirande, Dijon (France) and 2Unitb de Recherche Neuropharmacologique de I'INSERM, Paris (France) (Received 20 October 1987; Accepted 2 November 1987)
Key words." Lumbar locomotor generator; 2-Deoxyglucose; Rabbit An attempt was made to find the anatomical localization of the lumbar locomotion generators using 2-deoxy-[14C]glucose (2-DG) uptake in acute low spinal preparation of rabbits unanaesthetized, curarized and injected with nialamide and dihydroxyphenylalanine (DOPA). In such conditions, the locomotor generators were forced to work in isolation for 45 min without interruption as attested by the rhythmic activity recorded in hindlimb muscle nerves. Comlbared to spinal control preparations not activated pharmacologically, the treated animals showed a specific labeling in the intermediate part of the grey matter, extending from L6 to St.
It is well documented that the lumbar cord, deprived of supraspinal and peripheral informations, can generate locomotor-like activities, due to the presence of intrinsic generators [2, 7, 12]. During effective or fictive locomotor activity, lumbar neurones display bursty firing related to one phase of the ipsilateral locomotor cycle. These neurones are found in the intermediate and in the ventral part of the lumbar grey matter [3-5, 10]. Taking into account the dorsoventral sensorimotor organization of the spinal cord, one can assume that the more dorsally located bursty firing neurones are part of the lumbar locomotion generators. On the other hand, the rostrocaudal extent of such left and right generators is, up to date, not known. The present work aims to localize anatomically these generators using 2-deoxy-[14C]glucose (2-DG) uptake [11], a technique that allowed Batini et al. [1] to label the olivary generator responsible for harmaline tremor. We have used a preparation in which the lumbar generators, isolated from supraspinal levels and peripheral feedback are forced into continuous bursty firing by treatment with nialamide-dihydroxyphenylalanine (DOPA) [121. Seven young rabbits (500-1000 g body wt.) were used for the experiments. After Correspondence." D. Viala, Laboratoire de Neurophysiologie, UA C.N.R.S. 1199, Facult6 des Sciences Mirande, F-21004 Dijon, France. 0304-3940/88/$ 03.50 O 1988 Elsevier Scientific Publishers Ireland Ltd.
140
general anaesthesia with sodium pentobarbital (26 mg/kg i.v.) and tracheotomy, a catheter was inserted into one jugular vein for further injections. An extensive decortication was performed after craniectomy. Nerves to tibialis anterior (TA) and gastrocnemius medialis (GM) muscles were isolated and mounted on a pair of silver electrodes for recording of their efferent activities. A laminectomy was performed at T~2 level and the spinal cord was transected after local anaesthesia by lidocaine infiltration: the lumbar cord was thus isolated from supraspinal influences. The animals were then immobilized with flaxedil (to suppress peripheral feedback) and artificially ventilated. In 5 animals, nialamide (100 mg/kg i.v.) was injected just after general anaesthesia. M o t o r activities were recorded in hindlimb muscle nerves after administration of D O P A (100 mg/kg i.v.), 6 h after the onset of the experiment (this delay being necessary for the preparation to recover from general anaesthesia and for nialamide to become efficient). The efferent nerve activities were continuously monitored from 15 to 60 min (t~5 t60) after D O P A injection and stored in a tape recorder. The same protocol applies to the two other animals not injected with nialamide-DOPA and used as controls. All the animals received 100/~Ci/kg of [14C]2-DG i.v. at tls. They were sacrificed at t60 with a nembutal overdose. After extensive laminectomy, the lumbar cord was rapidly removed. Landmarks of vertebral limits were previously performed in order to determine thereafter the position of the spinal segments. The spinal segments L6 to $1 were taken off and frozen in isopentane precooled from - 4 5 to - 5 0 ° C . Serial sagittal or cross-sections (16/~m thick) were performed in a cryostat and were then rapidly dried at 40°C. The sections were exposed for autoradiography on X-ray film (single-coated Kodirex or single-coated Agfa Mamoray) for a few days. Thereafter the sections were stained for Nissl substance. In the control animals in which the lumbar locomotion generator was deprived of supraspinal and segmental control, the corresponding motoneurones were silent and no efferent activity could be disclosed in the hindlimb muscle nerves (Fig. I A). On the contrary, in the 5 experimental rabbits treated with nialamide-DOPA, continuous locomotor-like burst activity at the frequency of about 1 Hz [12] was observed in the muscle nerves throughout the 45 min of recording from tl5 to/60 (Fig. 1A'). The distribution of 2 - D G radioactivity was traced in the lumbosacral cords of the treated and of the control animals. The differences observed are shown in the crosssections of Fig. !B. In the control (a'), 3 zones of discrete labeling are met, due to different neurone density, corresponding to the substantia gelatinosa in the dorsal horn, to the central grey and to the motoneurone pools in the ventral horn (arrows in a and a'). The same regions of the cord are in general more labeled in treated animals. Moreover, there are two additional zones of specific labeling: one corresponding to the dorsal pool of motoneurones (horizontal arrow in b') and the second one corresponding to the intermediate part of the grey matter (coupled arrows in b'). Latter zone appears as a strip which extends from the dorsal pool of motoneurones (Mns) towards the medial border of the grey matter. Labeling of the dorsal pool of motoneurones is much more important than that of the ventral pool although both
141
Fig. 1. Labeling of the L7-S1 spinal cord in acute spinal preparations. A and A': recording of the spontaneous efferent activity in the tibialis anterior (TA) and gastrocnemius medialis (GM) muscle nerves in a control spinal preparation (A) and in another spinal preparation after nialamide-DOPA administration (A', 20 min after DOPA injection). In B, a,a' and b,b' show the labeled cross-sections at SI level (a and b) and their radioautography (a' and b') in a control preparation (a and a') and in a preparation injected with DOPA (b and b'). C: parasagittal slice of a spinal-DOPA preparation. Stained section (c) and its radioautography (c') are presented below each other. The vertical black line delimitates the substantia gelatinosa (its size is due to a torsion of this part of the cord) and the arrow indicates the maximal labeling in the intermediate grey matter.
142
dorsal and ventral Mns innervate hindlimb muscles and have been shown to be active during fictive spinal locomotion [8]. When observed with accuracy, the labeling in the region of the dorsal pool is not limited to the Mns but extends more medially. Therefore the labeling of this region is presumably due not only to the few motoneurones but also to other neighbouring neurones of the intermediate part of the cord. We have attempted to confirm this assumption and to determine the caudorostral extension of the labeling using sagittal serial sections performed at levels L 6 - S 1. In control preparations the findings were the following: in the lateral sections, light grey matter labeling corresponds to the substantia gelatinosa dorsally, and to the dorsal and ventral pools of motoneurones ventrally, these two structures being separated by almost unlabeled white matter; in the medial sections, light labeling is uniform and Mn label is undistinguishable from the surrounding. In the pharmacologically treated preparations, a general more important labeling of the grey matter is met. In addition, a specifically labeled zone is observed ventral to the substantia gelatinosa (latter delimited by a black vertical line on Fig. l c and c') and separated from it by an unlabeled band of 0.5 mm. This zone extends longitudinally from the caudal part of Sj to the caudal part of L 6 and it can be followed for 1.0 mm in the lateromedial direction. It clearly corresponds to the zone of intense labeling observed in the cross sections. The maximal density is at caudal level where the strip ends enlarged at a place where there is no histological substrate of increased neuronal density (wide arrow in Fig. 1Ccc'). These results show that the specific labeling does not correspond to the dorsal pool of Mns (the dorsal pool, which provides ankle and toe muscles, and the ventral pool that provides more proximal hindlimb muscles being equally labeled) but to an intermediate zone whose lateral edge is close to the dorsal pool of Mns. Because the maximal intensity of labeling corresponds to the intermediate grey matter where the more dorsal bursty firing neurones have been recorded during electrophysiological investigations (unpublished data), we may conclude that the neurones in this zone are active during locomotion and therefore belong to the central pattern generator, and that these active neurones have a higher metabolism than the follower neurones (including the Mns) during locomotion. The neurones which belong to the central pattern generator and the motoneurones are both active during locomotion. The reason why the generator neurones have a higher metabolism than the motoneurones cannot be answered with the present experiments: perhaps their density is higher or their firing level is much more increased than the motoneuronal one. Else this may possibly sign special metabolic requirements of these neurones if they are endogenous oscillators or pacemakers: this assumption cannot be excluded since some isolated neurones from the Mammalian spinal cord, whose function is still unknown, have been identified as endogenous oscillators [9]. Griilner [6] proposed an organization of the lumbar locomotion generator consisting in independent subgenerators coupled by interconnections. Our results are not in contradiction with this since the specifically labeled regions are not uniformly distributed within the caudo-rostral part of the cord. In addition, they suggest that the caudal part of the presumed generator (arrow in Fig. lCc') may play ~ special role of leader of the subgenerators.
143 We are grateful to Cesira Batini for her valuable comments
and suggestions on
this manuscript. 1 Batini, C., Buisseret-Delmas, C. and Conrath-Verrier, M., Harmaline induced tremor. I-Regional metabolic activity as revealed by ~4C-2-deoxyglucose in cat, Exp. Brain Res., 42 (1981) 371-382. 2 Brown, T.G., The intrinsic factors in the act of progression in the Mammal, Proc. R. Soc. Lond. B, 84 (1911) 308 319. 3 Edgerton, V.R., Grillner, S. and Sjostrom, A., On the spinal stepping generator, Soc. Neurosci. Abstr., I (1975) 615. 4 Edgerton, V.R., Grillner, S., Sjostrom, A. and Zangger, P., Central generation of locomotion in Vertebrates. In R.M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural Control of Locomotion, Vol. 18, New York, 1976, pp. 439~,64.. 5 Feldman, A.G. and Orlovsky, G.N., Activity of interneurones mediating reciprocal inhibition during locomotion, Brain Res., 84 (1975) 181 194. 6 Grillner, S., Control of locomotion in bipeds, tetrapods and fish. In J.M. Brookhart, V.B. Mountcastle, V.B. Brooks and S.R. Geiger (Eds.), Handbook of Physiology, Section 1: The Nervous System, Vol. II, Part 2, American Physiological Society, Bethesda, 1981, pp. 1179-1236. 7 Grillner, S. and Zangger, P., Locomotor movements generated by the deafferented spinal cord, Acta Physiol. Scand., 91 (1974) 38-39A. 8 Grillner, S. and Zangger, P., On the central generation of locomotion in the low spinal cat, Exp. Brain Res., 34 (1979) 241-261. 9 Legendre, P., McKenzie, J.S., Dupouy, B. and Vincent, J.D., Evidence for bursting pacemaker neurones in cultured spinal cord cells, Neuroscience, 16 (1985) 753 767. 10 Orlovsky, G.N. and Feldman, A.G., Classification of lumbo-sacral neurons according to their discharge patterns during evoked locomotion, Neurophysiology, 4 (1972) 410-417. 11 Sokoloff, L., Reivich, M., Kennedy, C., Des Rosiers, M.H., Patlak, C.S., Pettigrew, K.D., Sakurade, O. and Shinohara, M., The ~4C-deoxyglucose method for measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat, J. Neurochem., 28 (1977) 897 916. 12 Viala, D. and Buser, P., Modalit6s d'obtention de rythmes locomoteurs chez le lapin spinal par traitements pharmacologiques (DOPA, 5-HTP, D-amph&amine), Brain Res., 35 (I 971) 15 l - 165.