- Email: [email protected]
Maintenance of discrete somatosensory maps in subcortical relay nuclei is dependent on an intact sensory cortex
302
Developmental Brain Research, 44 ( 19881302-31 lg Flsevicr
BRD 6.0292
Maintenance of disc te somatceenso maps in cei relay nuclei is dependent on an intact sensory cortex R e h a S. E r z u r u m l u 1 a n d F . F . E b n e r 2 1Department of Brain and Cognitive Sciences, Whitaker College, Massachusetts Institute of Technology, Cambridge, MA 02139 (U.S.A.) and 2Centerfor Neural Science and Division of Biology and Medicine, Brown University, Providence, RI 02912 (U. S. A. )
(Accepted 23 August 1988) Key words: Somatosensorycortex; Barrel; Ventrobasal nucleus; Trigeminal system; Dorsal column nuclei; Pattern formation
Lesions of the rat barrelfield cortex drastically alter the discrete representations of the somatosensoryperiphery in the central nervous system. We have found that lesions placed in the parietal cortex, after the formation of barrels (postnatal day 5), can irreversibly abolish vibrissae- and extremity-related patterns of cytochrome oxidase activity in the principal sensory nucleus of the trigeminal nerve and in the dorsal column nuclei. Furthermore, abnormal patterns of enzymaticactivity occur in the remaining primary somatosensory cortex and the ventrobasal nucleus of the thalamus. We conclude that cortical barrels are essential in maintenance of periphery-related discrete morphologicalorganization in the rodent somatosensorysystem.
The rodent somatosensory system is characterized by discrete anatomical and physiological correlates of the sensory periphery. This organization was first noted, at the morphological level, by Woolsey and Van der Loos 28 in the mouse primary somatosensory (SI) cortex. They described a one-to-one relationship between the distribution of the vibrissae on the snout and the arrangement of cellular aggregates in layer IV of SI cortex, which they termed 'barrels'i Van der Loos and Woolsey25 further showed that during development, formation of barrels in the mouse SI cortex is dependent on an intact periphery and that perinatal damage to vibrissae follicles can irreversibly alter the organization of the barrels. Since then, numerous studies have explored the development and plasticity of the 'whisker-to-barrel' pathway in small rodents 1,3,4,6,9,14,15.24. It is now well established that the arrangement of the vibrissae on the snout and of the perioral sinus hairs is replicated in a punctate fashion first in the brainstem trigeminal complex (BSTC) 1'3'1°'19, then in the ventrobasal nucleus of the thalamus (VB) 3'23'29,
and finally in the posteromedial barrel subfield (PMBSF) of the SI cortex 15'22'24'2s. Furthermore, during development this organization proceeds sequentially from the periphery to the SI cortex 3'4'a'2~. Hence any disruption of the sensory periphery (by way of damage to the vibrissae follicles or to the infraorbital branch of the trigeminal nerve), at the time of translation of the pattern to the BSTC, irreversibly disrupts pattern formation in the VB nucleus and consequently in the SI cortex. These features of the trigeminal system are also shared by the dorsal column nuclear pathway to the SI cortex, where a discrete morphological pattern reflecting the distribution of tactile receptors of the paws is present 2"5. At all levels, these maps are easily discernible with cytochrome oxidase (CO) or succinic dehydrogenase (SDH) histochemistry 3'1s,19. Furthermore, there is ample evidence that high levels of oxidative metabolism, as assessed with SDH or CO histochemistry, correlates closely with the distribution of afferent terminal arbors at all levels of the rodent trigeminal systeml,15,18,19.
Correspondence: R.S. Erzurumlu, E25-634, Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, U.S.A.
0165-3806/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
303 Collectively these studies have arrived at the conclusion that a map of sensory receptor distribution in the periphery 'dictates' the formation of homeomorphic neural replicas in the central nervous system 24. The main question addressed in the present study is whether the periphery can still maintain the discrete morphological organization in subcortical somatosensory relay nuclei when the cortical barrels are disrupted. Here we present evidence that once the homeomorphic maps are established, their maintenance in the primary somatosensory relay nuclei is dependent on an intact barrelfield cortex. Long Evans hooded rats of both sexes (n = 14) were employed in this study. Unilateral lesions through the barrelfield cortex were performed on postnatal day 5 (day of birth: PND0), when barrels have just formed in the SI cortex4A5'2°. Since the trigeminal and the dorsal column nuclear pathways to the barrelfeld cortex are completely crossed, the contralateral somatosensory pathways served as controls for each animal. For surgery, pups were anesthetized with metolane on PND5 until they were immobile and insensitive to tail pinch. A mid-sagittal skin incision was made and a triangular bone flap was cut over the left hemisphere with fine scissors. The cortex was exposed by deflecting the bone flap rostrally at its apex. A square glass capillary tube (i.d. 0.5 mm), hydraulically coupled to a 10 ml syringe filled with Eagle's Basic Medium, was inserted through the neocortex in a caudolateral to rostromedial direction. Thus a rectangular strip of neocortex was isolated within the capillary. A shallow incision was made in the cortex at the tip of the capillary and the excised tissue was expelled by positive pressure on the fluid in the capillary. The extruded piece of cortex was picked up with gauze and discarded. The bone flap was pulled in place and the skin was sutured. Two postlesion survival time periods were used in this study. Since the discrete vibrissae- and paw-related patterns are best visualized with histochemical stains during the second week after birth, 12 animals were sacrifced on PND12 and two animals on PND40. The rats were anesthetized with an overdose of sodium pentobarbital and perfused transcardially first with heparinized saline followed by 4% paraformaldehyde in phosphate buffer (0.1 M, pH 7.4). The brains were removed, postfixed in the same fixative
overnight and cryoprotected with 20% sucrose in phosphate buffer. Frozen sections at a thickness of 100 p m were collected in phosphate buffer and reacted for CO activity27. The normal patterns of CO activity on PND12 in the SI cortex, VB nucleus, dorsal column nuclei and the 3 subdivisions of the BSTC are illustrated in Figs. I(A,C,E) and 2 (A,C,E), respectively. In each of these areas a discrete segmentation of CO activity pattern, related to the receptor distribution in the periphery is evident. A bilaminar pattern of high CO activity is seen in layers IV and VI of parietal cortex in the coronal plane. Barrels corresponding to the extremities are located medially while those related to the vibrissae are found laterally (Fig. 1A). Following lesions through the barrelfield on PND5, CO activity decreases in both layers of the remaining SI cortex (Fig. 1B). Further, no barrel-like segmentation is seen in layer IV but instead, a continuous band of lower CO activity is present (Fig. 1B). This fused band of low CO activity in layer IV persists at least until PND40 (Fig. 3B). The effects of barrelfeld lesions on the ipsilateral VB nucleus are illustrated in Fig. 1D. The VB nucleus undergoes rapid retrograde degeneration within one week after surgery and consequently, it is dramatically reduced in size. A rudimentary pattern of CO segmentation, with an abnormal organization, is often seen in the remaining VB nucleus (Fig. 1D). Normally, vibrissae-related barreloids form rows of arcs in the dorsomedial portion of the nucleus, while barreloids related to the extremities are located ventromedially and laterally (Fig. 1C). Within VB, there is a distinct border between the representation of the face and that of the rest of the body (separately named as the arcuate and the external divisions of the nucleus, respectively29). In the remnants of the VB nucleus ipsilateral to the cortical lesion, the border between the arcuate and the external divisions disappears. While the entire nucleus still retains its overall semicircular shape in the coronal plane, abnormal patches of CO activity are seen within it (Fig. 1D). Following barrelfield cortex lesions, dramatic changes in the pattern of CO activity are seen in the contralateral somatosensory brainstem. In the dorsal column nuclei, normally the most prominent pattern seen is in the cuneate nucleus where palm pads and digits of the forepaw are represented as discrete, CO-
304
Fig. 1. Normal CO activity pattern on PNDI2 in barrelfield cortex (A), VB (C) and dorsal column nuclei (E) and the altered pattern of CO activity in these areas (B, D and F). In A, medially placed arrowhead points to the barrels related to the forepaw and laterally placed arrowhead to the vibrissae-related barrels. Note that in the lesioned side the remaining SI cortex (B) has a fused band of low CO activity in layer IV (arrowhead). In C the border between the medially situated arcuate and the laterally situated external divisions of the VB is denoted by asterisks. Note that the VB on the lesioned side is considerably smaller in size, with abnormal patches of CO activity and no distinct border between the arcuate and the external divisions. The normal pattern of CO activity in cuneate (C) and gracile (G) nuclei is shown in E. Contralateral to the lesioned barrelfield cortex these nuclei undergo drastic changes in their CO activity pattern (micrograph F). Bar = 1.5 mm (A,B); 0.53 mm (C.D) and 0.38 mm rE,F). All micrographs arc taken from thc same case.
305
E Fig. 2. Brainstem trigeminal complex. The micrographs on the left (A, C and E) show the normal pattern of vibrissae-related CO activity in the principal sensory nucleus (A), subnucleus interpolaris (C) and subnucleus caudalis (E). In all these areas representation of 5 rows of whiskers (rows A - E ) are denoted. The micrographs on the right side of the figure illustrate the differential effects of PND5 barrelfield lesions on the contralateral BSTC of the same animal. Note that while the patterns in subnuclei interpolaris (D) and caudalis (F) are normal, the principal sensory nucleus totally lacks any vibrissae-related pattern (B). Bar = 0.5 mm (A-F).
306
~iii~¸¸
~iili~~ ~
i ¸¸
II
ii:!i !iiiii~ ~!~i~
II
I
I
Fig. 3. CO activity patterns m the normal (A) and lesioned (B) cortex and the related principal sensory nuclei (C and D, respectively) of an animal that underwent barrelfield lesion on PND5 and was sacrificed on PND40, For better comparison, both principal sensory nuclei are presented in the same orientation. Bar = 1.3 mm (A) and t).38 mm (D).
dense patches (Fig. 1E). Within a week after cortical lesions the segmentation in the contralateral cuneate nucleus is totally abolished and the overall C O activity is r e d u c e d (Fig. IF). Similar effects are also seen in the gracile nucleus in the region of the hindpaw representation.
In the BSTC, trigeminal afferents are distributed in the principal sensory nucleus and all 3 subdivisions of the spinal trigeminal nucleus (rostrocaudally: subnuclei oralis, interpolaris and caudalis) 1. H o w e v e r , vibrissae-related segmentation is normally not seen in subnucleus oralis with C O or S D H histochemistry,
307 while it is distinct in the principal sensory nucleus and the subnuclei interpolaris and caudalis of the spinal trigeminal nucleus (Fig. 2A,C,E). Of these 3 areas with distinct patterns of CO activity, only the principal sensory nucleus is responsive to damage of the contralateral barrelfield cortex. Normally in the ventral portion of the principal sensory nucleus, 5 rows of vibrissae-related segmentation are seen (Fig. 2A). Following lesions of the contralateral barrelfield cortex, the density of CO activity within the nucleus decreases and the segmentation disappears altogether (Fig. 2B). In contrast, vibrissae-related patterns in subnuclei interpolaris and caudalis are not affected by such lesions (Fig. 2D,F). The disruption of the pattern in the principal sensory nucleus and in the dorsal column nuclei appears to be irreversible, as it persists at least as late as PND40 (Fig. 3D). The results of the present study reveal that somatosensory maps in the primary relay nuclei (namely, the principal sensory nucleus of the trigeminal nerve and the dorsal column nuclei) are not exclusively dependent on the periphery. Previous studies on the development and modifiability of the somatosensory pathways have emphasized the role of the sensory periphery in providing the 'blueprint' for the discrete maps in the brain 3'5'9A3'24'25. Our results provide evidence that, once established, these maps engage in a mutually dependent interaction between different regions of the somatosensory neuraxis. In a parallel study, Ito and Seo 12'21 made focal lesions of the developing barrelfield cortex on different postnatal days (PNDs 0, 5, 14 and 20) and assessed the effects of such lesions on the integrity of CO and SDH activity patterns in the adult barrelfield cortex. They observed that such small lesions are circumvented by the developing barrels and a full set of barrels is maintained if the lesions are made on or before PND10. They reported that later lesions (on PNDs 14 and 20) are contained within the barrelfield with no apparent distortion of the remaining barrels. At the cortical level, the differences in the outcome of their study and ours may be attributed to the size of the lesions placed in the cortex. While in their study the lesions span at most the width of a few barrels in layer IV, in the present study the lesions were large enough to go through all layers of the SI cortex and across rows of barrels. Furthermore it is important to note that such large lesions can extensively perturb
specific thalamocortical connectivity and consequently abolish already formed barrels in the cortex. While the effects of cortical lesions on the VB nucleus can be explained by retrograde degeneration and abnormal clumping of lemniscal afferents in the remaining nuclear area, the disappearance of the periphery-related patterns in the brainstem is intriguing. The distribution of the primary sensory afferent arbors in the somatosensory brainstem nuclei closely correlates with SDH or CO segmentation 1. Hence, the effects on the principal sensory nucleus and the dorsal column nuclei cannot be solely attributed to retrograde and transneuronal degeneration. The trigeminal and dorsal root ganglion afferents are 3 synapses away from the lesion site, yet their normal CO activity pattern is abolished within a week. It is also unlikely that disruption of cortical projections from layer Vb to these somatosensory brainstem structures 26 is responsible for the disappearance of the discrete patterns, because the patterns in the spinal trigeminal nucleus are not affected. The main projection target of both the principal sensory nucleus and the dorsal column nuclei is the VB nucleus, while the subnucleus interpolaris has other substantial projections to the cerebellum, inferior olive, superior colliculus and zona incerta 7'1m6. Similarly, the magnocellular portion of the subnucleus caudalis, which contains the vibrissae-related segmentation, projects primarily to the facial nucleuss. It has been shown that the principal sensory nucleus alone provides the template for the vibrissaerelated pattern to the VB nucleus and consequently to the SI cortex 17. Furthermore, the relay neurons of the principal sensory nucleus are also organized into 'barreloids' while those of the subnucleus interpolaris are not a°. Such differences in projection patterns and cytoarchitecture between the principal sensory nucleus and the spinal trigeminal nucleus provide the simplest explanation for the differential effects of the cortical lesions on the BSTC. In addition, these results support the contention that somatosensory pathways diverge in the brainstem and spinal cord levels: one pathway operating through the principal sensory nucleus-dorsal column nuclei-VB nucleusbarrelfield cortex and the other through the spinal trigeminal nucleus and dorsal horns at subcortical levels. We conclude that in the primary somatosensory
308 p a t h w a y discrete m a p s of the sensory p e r i p h e r y are
W e wish to t h a n k Drs. S. J h a v e r i and H . P . Killack-
m u t u a l l y d e p e n d e n t on o n e a n o t h e r for their integri-
ey for their helpful suggestions and criticisms. Re-
ty and that cortical d a m a g e can alter these m a p s e v e n
search s u p p o r t e d by M a t h e r s F o u n d a t i o n .
in the p r e s e n c e of an intact p e r i p h e r y .
1 Bates, C.A. and Killackey, H.P., The organization of the neonatal rat's brainstem trigeminal complex and its role in the formation of central trigeminal patterns, J. Comp. Neurol., 240 (1985) 513-538. 2 Belford, G.R. and Killackey, H.P., Anatomical correlates of the forelimb in the ventrobasal complex and the cuneate nucleus of the neontal rat, Brain Res., 158 (1978) 450-455. 3 Belford, G.R. and Killackey, H.P., Vibrissae representation in subcortical trigeminal centers of the neonatal rat, J. Comp. Neurol., 183 (1979) 305-322. 4 Belford, G.R. and Killackey, H.P., The sensitive period in the development of the trigeminal system of the neonatal rat, J. Comp. NeuroL, 229 (1980) 199-213. 5 Dawson, D.R. and KiUackey, H.P., The organization and mutability of the forepaw and hindpaw representations in the somatosensory cortex of the neonatal rat, J. Comp. Neurol., 256 (1987) 246-256. 6 Durham, D. and Woolsey, T.A., Effects of neonatal whisker lesions on mouse central trigeminal pathways, J. Comp. NeuroL, 223 (1984) 424-447. 7 Erzurumlu, R.S. and Killackey, H.P., Efferent connections of the brainstem trigeminal complex with the facial nucleus of the rat, J. Comp. Neurol., 188 (1979) 75-86. 8 Erzurumlu, R.S. and Killackey, H.P., Diencephalic projections of the subnucleus interpolaris of the brainstem trigeminal complex in the rat, Neuroscience, 5 (1980) 1891-1901. 9 Erzurumlu, R.S. and Killackey, H.P., Development of order in the rat trigeminal system, J. Comp. Neurol., 213 (1983) 365-380. 10 Erzurumlu, R.S., Bates, C.A. and Killackey, H.P., Differential organization of thalamic projection cells in the brainstem trigeminal complex of the rat, Brain Res., 198 (1980) 427-433. 11 Huerta, M.F., Frankfurter, A. and Harting, J.K., Studies of the principal sensory and spinal trigeminal nuclei of the rat: projections to the superior colliculus, inferior olive, and cerebellum, J. Comp. Neurol., 220 (1983) 147-167. 12 Ito, M. and Seo, M.L., Avoidance of neonatal cortical lesions by developing somatosensory barrels, Nature (Lond.), 301 (1983) 600-602. 13 Jeanmonod, D., Rice, F.L. and Van der Loos, H., Mouse somatosensory cortex: alterations in the barrelfield following receptor injury at different early postnatal ages, Neuroscience, 6 (1981) 1503-1535. 14 Killackey, H.P., Pattern formation in the trigeminal system of the rat, Trends Neurosci., 3 (1980) 303-306. 15 Killackey, H.P. and Belford, G.R., The formation of afferent patterns in the somatosensory cortex of the neonatal
rat, J. Comp. Neurol., 183 (1979) 285-304. 16 Killackey, H.P. and Erzurumlu, R.S., Trigeminal projections to the superior colliculus of the rat, J. Comp. Neurol., 201 (1981) 221-242. 17 Killackey, H.P. and Fleming, K., The role of the principal sensory nucleus in central trigeminal pattern formation, Dev. Brain Res., 22 (1985) 141-145. 18 Land, P.W. and Simons, D.J., Cytochrome oxidase staining in the rat SmI barrel cortex, J. Comp. Neurol., 238 (1985) 225-235. 19 Ma, P.M. and Woolsey, T.A., Cytoarchitectonic correlates of the vibrissae in the medullary trigeminal complex of the mouse, Brain Res., 306 (1984) 374-379. 20 Rice, F.L. and Van der Loos, H., Development of the barrels and barrel field in the somatosensory cortex of the mouse,J. Comp. Neurol., 171 (1977)545-560. 21 Seo, M.L. and Ito, M., Reorganization of rat vibrissa barrelfield as studied by cortical lesioning on different postnatal days, Exp. Brain Res., 65 (1987) 251-260. 22 Simons, D.J., Durham, D. and Woolsey, T.A., Functional organization of mouse and rat SmI barrel cortex following vibrissal damage on different postnatal days, Somatosens. Res., 1 (1984) 207-245. 23 Van der Loos, H., Barreloids in mouse somatosensory thalamus, Neurosci. Lett., 2 (1976) 1-6. 24 Van der Loos, H. and Welker, E., Development and plasticity of somatosensory brain maps. In M. Rowe and W.D. Willis (Eds.), Development, Organization and Processing in Somatosensory Pathways, Liss, New York, 1985, pp. 53-67. 25 Van der Loos, H. and Woolsey, T.A., Somatosensory cortex: structural alterations following early injury to sense organs, Science, 179 (1973) 395-398. 26 Wise, S.P. and Jones, E.G., Cells of origin and terminal distribution of descending projections of the rat somatic sensory cortex, J. Comp. Neurol., 175 (1977) 129-158. 27 Wong-Riley, M.T.T., Changes in the visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry, Brain Res., 171 (1979) 11-28. 28 Woolsey, T.A. and Van der Loos, H., The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex, Brain Res., 17 (1970) 205-242. 29 Woolsey, T.A., Anderson, J.R., Wann, J.R. and Stanfield, B.B., Effects of early vibrissae damage on neurons in the ventrobasal (VB) thalamus of the mouse, J. Comp. Neurol., 184 (1979) 363-380.
Developmental Brain Research, 44 ( 19881302-31 lg Flsevicr
BRD 6.0292
Maintenance of disc te somatceenso maps in cei relay nuclei is dependent on an intact sensory cortex R e h a S. E r z u r u m l u 1 a n d F . F . E b n e r 2 1Department of Brain and Cognitive Sciences, Whitaker College, Massachusetts Institute of Technology, Cambridge, MA 02139 (U.S.A.) and 2Centerfor Neural Science and Division of Biology and Medicine, Brown University, Providence, RI 02912 (U. S. A. )
(Accepted 23 August 1988) Key words: Somatosensorycortex; Barrel; Ventrobasal nucleus; Trigeminal system; Dorsal column nuclei; Pattern formation
Lesions of the rat barrelfield cortex drastically alter the discrete representations of the somatosensoryperiphery in the central nervous system. We have found that lesions placed in the parietal cortex, after the formation of barrels (postnatal day 5), can irreversibly abolish vibrissae- and extremity-related patterns of cytochrome oxidase activity in the principal sensory nucleus of the trigeminal nerve and in the dorsal column nuclei. Furthermore, abnormal patterns of enzymaticactivity occur in the remaining primary somatosensory cortex and the ventrobasal nucleus of the thalamus. We conclude that cortical barrels are essential in maintenance of periphery-related discrete morphologicalorganization in the rodent somatosensorysystem.
The rodent somatosensory system is characterized by discrete anatomical and physiological correlates of the sensory periphery. This organization was first noted, at the morphological level, by Woolsey and Van der Loos 28 in the mouse primary somatosensory (SI) cortex. They described a one-to-one relationship between the distribution of the vibrissae on the snout and the arrangement of cellular aggregates in layer IV of SI cortex, which they termed 'barrels'i Van der Loos and Woolsey25 further showed that during development, formation of barrels in the mouse SI cortex is dependent on an intact periphery and that perinatal damage to vibrissae follicles can irreversibly alter the organization of the barrels. Since then, numerous studies have explored the development and plasticity of the 'whisker-to-barrel' pathway in small rodents 1,3,4,6,9,14,15.24. It is now well established that the arrangement of the vibrissae on the snout and of the perioral sinus hairs is replicated in a punctate fashion first in the brainstem trigeminal complex (BSTC) 1'3'1°'19, then in the ventrobasal nucleus of the thalamus (VB) 3'23'29,
and finally in the posteromedial barrel subfield (PMBSF) of the SI cortex 15'22'24'2s. Furthermore, during development this organization proceeds sequentially from the periphery to the SI cortex 3'4'a'2~. Hence any disruption of the sensory periphery (by way of damage to the vibrissae follicles or to the infraorbital branch of the trigeminal nerve), at the time of translation of the pattern to the BSTC, irreversibly disrupts pattern formation in the VB nucleus and consequently in the SI cortex. These features of the trigeminal system are also shared by the dorsal column nuclear pathway to the SI cortex, where a discrete morphological pattern reflecting the distribution of tactile receptors of the paws is present 2"5. At all levels, these maps are easily discernible with cytochrome oxidase (CO) or succinic dehydrogenase (SDH) histochemistry 3'1s,19. Furthermore, there is ample evidence that high levels of oxidative metabolism, as assessed with SDH or CO histochemistry, correlates closely with the distribution of afferent terminal arbors at all levels of the rodent trigeminal systeml,15,18,19.
Correspondence: R.S. Erzurumlu, E25-634, Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, U.S.A.
0165-3806/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
303 Collectively these studies have arrived at the conclusion that a map of sensory receptor distribution in the periphery 'dictates' the formation of homeomorphic neural replicas in the central nervous system 24. The main question addressed in the present study is whether the periphery can still maintain the discrete morphological organization in subcortical somatosensory relay nuclei when the cortical barrels are disrupted. Here we present evidence that once the homeomorphic maps are established, their maintenance in the primary somatosensory relay nuclei is dependent on an intact barrelfield cortex. Long Evans hooded rats of both sexes (n = 14) were employed in this study. Unilateral lesions through the barrelfield cortex were performed on postnatal day 5 (day of birth: PND0), when barrels have just formed in the SI cortex4A5'2°. Since the trigeminal and the dorsal column nuclear pathways to the barrelfeld cortex are completely crossed, the contralateral somatosensory pathways served as controls for each animal. For surgery, pups were anesthetized with metolane on PND5 until they were immobile and insensitive to tail pinch. A mid-sagittal skin incision was made and a triangular bone flap was cut over the left hemisphere with fine scissors. The cortex was exposed by deflecting the bone flap rostrally at its apex. A square glass capillary tube (i.d. 0.5 mm), hydraulically coupled to a 10 ml syringe filled with Eagle's Basic Medium, was inserted through the neocortex in a caudolateral to rostromedial direction. Thus a rectangular strip of neocortex was isolated within the capillary. A shallow incision was made in the cortex at the tip of the capillary and the excised tissue was expelled by positive pressure on the fluid in the capillary. The extruded piece of cortex was picked up with gauze and discarded. The bone flap was pulled in place and the skin was sutured. Two postlesion survival time periods were used in this study. Since the discrete vibrissae- and paw-related patterns are best visualized with histochemical stains during the second week after birth, 12 animals were sacrifced on PND12 and two animals on PND40. The rats were anesthetized with an overdose of sodium pentobarbital and perfused transcardially first with heparinized saline followed by 4% paraformaldehyde in phosphate buffer (0.1 M, pH 7.4). The brains were removed, postfixed in the same fixative
overnight and cryoprotected with 20% sucrose in phosphate buffer. Frozen sections at a thickness of 100 p m were collected in phosphate buffer and reacted for CO activity27. The normal patterns of CO activity on PND12 in the SI cortex, VB nucleus, dorsal column nuclei and the 3 subdivisions of the BSTC are illustrated in Figs. I(A,C,E) and 2 (A,C,E), respectively. In each of these areas a discrete segmentation of CO activity pattern, related to the receptor distribution in the periphery is evident. A bilaminar pattern of high CO activity is seen in layers IV and VI of parietal cortex in the coronal plane. Barrels corresponding to the extremities are located medially while those related to the vibrissae are found laterally (Fig. 1A). Following lesions through the barrelfield on PND5, CO activity decreases in both layers of the remaining SI cortex (Fig. 1B). Further, no barrel-like segmentation is seen in layer IV but instead, a continuous band of lower CO activity is present (Fig. 1B). This fused band of low CO activity in layer IV persists at least until PND40 (Fig. 3B). The effects of barrelfeld lesions on the ipsilateral VB nucleus are illustrated in Fig. 1D. The VB nucleus undergoes rapid retrograde degeneration within one week after surgery and consequently, it is dramatically reduced in size. A rudimentary pattern of CO segmentation, with an abnormal organization, is often seen in the remaining VB nucleus (Fig. 1D). Normally, vibrissae-related barreloids form rows of arcs in the dorsomedial portion of the nucleus, while barreloids related to the extremities are located ventromedially and laterally (Fig. 1C). Within VB, there is a distinct border between the representation of the face and that of the rest of the body (separately named as the arcuate and the external divisions of the nucleus, respectively29). In the remnants of the VB nucleus ipsilateral to the cortical lesion, the border between the arcuate and the external divisions disappears. While the entire nucleus still retains its overall semicircular shape in the coronal plane, abnormal patches of CO activity are seen within it (Fig. 1D). Following barrelfield cortex lesions, dramatic changes in the pattern of CO activity are seen in the contralateral somatosensory brainstem. In the dorsal column nuclei, normally the most prominent pattern seen is in the cuneate nucleus where palm pads and digits of the forepaw are represented as discrete, CO-
304
Fig. 1. Normal CO activity pattern on PNDI2 in barrelfield cortex (A), VB (C) and dorsal column nuclei (E) and the altered pattern of CO activity in these areas (B, D and F). In A, medially placed arrowhead points to the barrels related to the forepaw and laterally placed arrowhead to the vibrissae-related barrels. Note that in the lesioned side the remaining SI cortex (B) has a fused band of low CO activity in layer IV (arrowhead). In C the border between the medially situated arcuate and the laterally situated external divisions of the VB is denoted by asterisks. Note that the VB on the lesioned side is considerably smaller in size, with abnormal patches of CO activity and no distinct border between the arcuate and the external divisions. The normal pattern of CO activity in cuneate (C) and gracile (G) nuclei is shown in E. Contralateral to the lesioned barrelfield cortex these nuclei undergo drastic changes in their CO activity pattern (micrograph F). Bar = 1.5 mm (A,B); 0.53 mm (C.D) and 0.38 mm rE,F). All micrographs arc taken from thc same case.
305
E Fig. 2. Brainstem trigeminal complex. The micrographs on the left (A, C and E) show the normal pattern of vibrissae-related CO activity in the principal sensory nucleus (A), subnucleus interpolaris (C) and subnucleus caudalis (E). In all these areas representation of 5 rows of whiskers (rows A - E ) are denoted. The micrographs on the right side of the figure illustrate the differential effects of PND5 barrelfield lesions on the contralateral BSTC of the same animal. Note that while the patterns in subnuclei interpolaris (D) and caudalis (F) are normal, the principal sensory nucleus totally lacks any vibrissae-related pattern (B). Bar = 0.5 mm (A-F).
306
~iii~¸¸
~iili~~ ~
i ¸¸
II
ii:!i !iiiii~ ~!~i~
II
I
I
Fig. 3. CO activity patterns m the normal (A) and lesioned (B) cortex and the related principal sensory nuclei (C and D, respectively) of an animal that underwent barrelfield lesion on PND5 and was sacrificed on PND40, For better comparison, both principal sensory nuclei are presented in the same orientation. Bar = 1.3 mm (A) and t).38 mm (D).
dense patches (Fig. 1E). Within a week after cortical lesions the segmentation in the contralateral cuneate nucleus is totally abolished and the overall C O activity is r e d u c e d (Fig. IF). Similar effects are also seen in the gracile nucleus in the region of the hindpaw representation.
In the BSTC, trigeminal afferents are distributed in the principal sensory nucleus and all 3 subdivisions of the spinal trigeminal nucleus (rostrocaudally: subnuclei oralis, interpolaris and caudalis) 1. H o w e v e r , vibrissae-related segmentation is normally not seen in subnucleus oralis with C O or S D H histochemistry,
307 while it is distinct in the principal sensory nucleus and the subnuclei interpolaris and caudalis of the spinal trigeminal nucleus (Fig. 2A,C,E). Of these 3 areas with distinct patterns of CO activity, only the principal sensory nucleus is responsive to damage of the contralateral barrelfield cortex. Normally in the ventral portion of the principal sensory nucleus, 5 rows of vibrissae-related segmentation are seen (Fig. 2A). Following lesions of the contralateral barrelfield cortex, the density of CO activity within the nucleus decreases and the segmentation disappears altogether (Fig. 2B). In contrast, vibrissae-related patterns in subnuclei interpolaris and caudalis are not affected by such lesions (Fig. 2D,F). The disruption of the pattern in the principal sensory nucleus and in the dorsal column nuclei appears to be irreversible, as it persists at least as late as PND40 (Fig. 3D). The results of the present study reveal that somatosensory maps in the primary relay nuclei (namely, the principal sensory nucleus of the trigeminal nerve and the dorsal column nuclei) are not exclusively dependent on the periphery. Previous studies on the development and modifiability of the somatosensory pathways have emphasized the role of the sensory periphery in providing the 'blueprint' for the discrete maps in the brain 3'5'9A3'24'25. Our results provide evidence that, once established, these maps engage in a mutually dependent interaction between different regions of the somatosensory neuraxis. In a parallel study, Ito and Seo 12'21 made focal lesions of the developing barrelfield cortex on different postnatal days (PNDs 0, 5, 14 and 20) and assessed the effects of such lesions on the integrity of CO and SDH activity patterns in the adult barrelfield cortex. They observed that such small lesions are circumvented by the developing barrels and a full set of barrels is maintained if the lesions are made on or before PND10. They reported that later lesions (on PNDs 14 and 20) are contained within the barrelfield with no apparent distortion of the remaining barrels. At the cortical level, the differences in the outcome of their study and ours may be attributed to the size of the lesions placed in the cortex. While in their study the lesions span at most the width of a few barrels in layer IV, in the present study the lesions were large enough to go through all layers of the SI cortex and across rows of barrels. Furthermore it is important to note that such large lesions can extensively perturb
specific thalamocortical connectivity and consequently abolish already formed barrels in the cortex. While the effects of cortical lesions on the VB nucleus can be explained by retrograde degeneration and abnormal clumping of lemniscal afferents in the remaining nuclear area, the disappearance of the periphery-related patterns in the brainstem is intriguing. The distribution of the primary sensory afferent arbors in the somatosensory brainstem nuclei closely correlates with SDH or CO segmentation 1. Hence, the effects on the principal sensory nucleus and the dorsal column nuclei cannot be solely attributed to retrograde and transneuronal degeneration. The trigeminal and dorsal root ganglion afferents are 3 synapses away from the lesion site, yet their normal CO activity pattern is abolished within a week. It is also unlikely that disruption of cortical projections from layer Vb to these somatosensory brainstem structures 26 is responsible for the disappearance of the discrete patterns, because the patterns in the spinal trigeminal nucleus are not affected. The main projection target of both the principal sensory nucleus and the dorsal column nuclei is the VB nucleus, while the subnucleus interpolaris has other substantial projections to the cerebellum, inferior olive, superior colliculus and zona incerta 7'1m6. Similarly, the magnocellular portion of the subnucleus caudalis, which contains the vibrissae-related segmentation, projects primarily to the facial nucleuss. It has been shown that the principal sensory nucleus alone provides the template for the vibrissaerelated pattern to the VB nucleus and consequently to the SI cortex 17. Furthermore, the relay neurons of the principal sensory nucleus are also organized into 'barreloids' while those of the subnucleus interpolaris are not a°. Such differences in projection patterns and cytoarchitecture between the principal sensory nucleus and the spinal trigeminal nucleus provide the simplest explanation for the differential effects of the cortical lesions on the BSTC. In addition, these results support the contention that somatosensory pathways diverge in the brainstem and spinal cord levels: one pathway operating through the principal sensory nucleus-dorsal column nuclei-VB nucleusbarrelfield cortex and the other through the spinal trigeminal nucleus and dorsal horns at subcortical levels. We conclude that in the primary somatosensory
308 p a t h w a y discrete m a p s of the sensory p e r i p h e r y are
W e wish to t h a n k Drs. S. J h a v e r i and H . P . Killack-
m u t u a l l y d e p e n d e n t on o n e a n o t h e r for their integri-
ey for their helpful suggestions and criticisms. Re-
ty and that cortical d a m a g e can alter these m a p s e v e n
search s u p p o r t e d by M a t h e r s F o u n d a t i o n .
in the p r e s e n c e of an intact p e r i p h e r y .
1 Bates, C.A. and Killackey, H.P., The organization of the neonatal rat's brainstem trigeminal complex and its role in the formation of central trigeminal patterns, J. Comp. Neurol., 240 (1985) 513-538. 2 Belford, G.R. and Killackey, H.P., Anatomical correlates of the forelimb in the ventrobasal complex and the cuneate nucleus of the neontal rat, Brain Res., 158 (1978) 450-455. 3 Belford, G.R. and Killackey, H.P., Vibrissae representation in subcortical trigeminal centers of the neonatal rat, J. Comp. Neurol., 183 (1979) 305-322. 4 Belford, G.R. and Killackey, H.P., The sensitive period in the development of the trigeminal system of the neonatal rat, J. Comp. NeuroL, 229 (1980) 199-213. 5 Dawson, D.R. and KiUackey, H.P., The organization and mutability of the forepaw and hindpaw representations in the somatosensory cortex of the neonatal rat, J. Comp. Neurol., 256 (1987) 246-256. 6 Durham, D. and Woolsey, T.A., Effects of neonatal whisker lesions on mouse central trigeminal pathways, J. Comp. NeuroL, 223 (1984) 424-447. 7 Erzurumlu, R.S. and Killackey, H.P., Efferent connections of the brainstem trigeminal complex with the facial nucleus of the rat, J. Comp. Neurol., 188 (1979) 75-86. 8 Erzurumlu, R.S. and Killackey, H.P., Diencephalic projections of the subnucleus interpolaris of the brainstem trigeminal complex in the rat, Neuroscience, 5 (1980) 1891-1901. 9 Erzurumlu, R.S. and Killackey, H.P., Development of order in the rat trigeminal system, J. Comp. Neurol., 213 (1983) 365-380. 10 Erzurumlu, R.S., Bates, C.A. and Killackey, H.P., Differential organization of thalamic projection cells in the brainstem trigeminal complex of the rat, Brain Res., 198 (1980) 427-433. 11 Huerta, M.F., Frankfurter, A. and Harting, J.K., Studies of the principal sensory and spinal trigeminal nuclei of the rat: projections to the superior colliculus, inferior olive, and cerebellum, J. Comp. Neurol., 220 (1983) 147-167. 12 Ito, M. and Seo, M.L., Avoidance of neonatal cortical lesions by developing somatosensory barrels, Nature (Lond.), 301 (1983) 600-602. 13 Jeanmonod, D., Rice, F.L. and Van der Loos, H., Mouse somatosensory cortex: alterations in the barrelfield following receptor injury at different early postnatal ages, Neuroscience, 6 (1981) 1503-1535. 14 Killackey, H.P., Pattern formation in the trigeminal system of the rat, Trends Neurosci., 3 (1980) 303-306. 15 Killackey, H.P. and Belford, G.R., The formation of afferent patterns in the somatosensory cortex of the neonatal
rat, J. Comp. Neurol., 183 (1979) 285-304. 16 Killackey, H.P. and Erzurumlu, R.S., Trigeminal projections to the superior colliculus of the rat, J. Comp. Neurol., 201 (1981) 221-242. 17 Killackey, H.P. and Fleming, K., The role of the principal sensory nucleus in central trigeminal pattern formation, Dev. Brain Res., 22 (1985) 141-145. 18 Land, P.W. and Simons, D.J., Cytochrome oxidase staining in the rat SmI barrel cortex, J. Comp. Neurol., 238 (1985) 225-235. 19 Ma, P.M. and Woolsey, T.A., Cytoarchitectonic correlates of the vibrissae in the medullary trigeminal complex of the mouse, Brain Res., 306 (1984) 374-379. 20 Rice, F.L. and Van der Loos, H., Development of the barrels and barrel field in the somatosensory cortex of the mouse,J. Comp. Neurol., 171 (1977)545-560. 21 Seo, M.L. and Ito, M., Reorganization of rat vibrissa barrelfield as studied by cortical lesioning on different postnatal days, Exp. Brain Res., 65 (1987) 251-260. 22 Simons, D.J., Durham, D. and Woolsey, T.A., Functional organization of mouse and rat SmI barrel cortex following vibrissal damage on different postnatal days, Somatosens. Res., 1 (1984) 207-245. 23 Van der Loos, H., Barreloids in mouse somatosensory thalamus, Neurosci. Lett., 2 (1976) 1-6. 24 Van der Loos, H. and Welker, E., Development and plasticity of somatosensory brain maps. In M. Rowe and W.D. Willis (Eds.), Development, Organization and Processing in Somatosensory Pathways, Liss, New York, 1985, pp. 53-67. 25 Van der Loos, H. and Woolsey, T.A., Somatosensory cortex: structural alterations following early injury to sense organs, Science, 179 (1973) 395-398. 26 Wise, S.P. and Jones, E.G., Cells of origin and terminal distribution of descending projections of the rat somatic sensory cortex, J. Comp. Neurol., 175 (1977) 129-158. 27 Wong-Riley, M.T.T., Changes in the visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry, Brain Res., 171 (1979) 11-28. 28 Woolsey, T.A. and Van der Loos, H., The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex, Brain Res., 17 (1970) 205-242. 29 Woolsey, T.A., Anderson, J.R., Wann, J.R. and Stanfield, B.B., Effects of early vibrissae damage on neurons in the ventrobasal (VB) thalamus of the mouse, J. Comp. Neurol., 184 (1979) 363-380.