EXPERIMENTAL
115,121-126
NEUROLOGY
(1992)
The Specification of Sensory Cortex: Lessons from Cortical Transplantation DENNISD.M.O'LEARY,*BRADLEYL.SCHLAGGAR,*ANDBRENTB.STANFIELD~ *Molecular
Neurobiology Laboratory, of Mental Health,
The Sadk Institute, La Jolla, California 92037; and tNationa1 NIH Animal Center, Poolesuille, Maryland 92837
Institute
programs and environmental influences to the phenomenon of interest. In one extreme, cortical areas may be entirely prespecified-the ultimate emergence of areaspecific features is a programmed event specified in the neuroepithelium that gives rise to the cortex. In the other extreme, area-specific features may be entirely imposed upon a region of developing cortex by outside influences-the distinctive characteristics of an area result from interactions between the already assembled, yet immature cortex and agents extrinsic to the cortex. Of course, neither of these extremes is likely to be entirely correct, but they help to focus attention on an issue antecedent to an understanding of the mechanisms involved in the emergence of distinct cortical areas. Over the past several years, we have attempted to determine the extent to which the neocortical neuroepithelium is regionally specified to generate area-unique cohorts of cells programmed to develop area-specific features or, conversely, is lacking such regionalization and generates comparable populations of cells across its extent. Specifically, we have employed a heterotopic transplantation paradigm to test the potential of regions of cortex to differentiate features that are normally unique to other areas (19, 23, 30, 32). The transplantation of developing tissue, a method of great historical significance in experimental embryology and developmental neurobiology, is a powerful tool for addressing such questions of fate and potential (8). Several researchers, especially Lund and his colleagues, have demonstrated that fetal rat neocortex can be successfully transplanted: transplanted embryonic rat cortex continues to generate cells, develops a laminar structure, and forms connections with the host brain (3, 4, 11).
The mammalian neocortex is functionally organized into numerous specia!ized “areas.” The distinct functional properties characteristic of each area are in large part due to connectional and architectural differences among the areas. However, these “area-specific” features which distinguish mature areas are not apparent early in the development of neocortex. We have used heterotopic cortical transplantation to examine whether these area-specific features are prespecified or emerge as a result of epigenetic interactions. Here, we review our studies in which late fetal rat cortex was transplanted heterotopically into the cortex of newborn rats to test its capacity to differentiate features normally unique to other cortical areas. We find that regions of the developing neocortex have similar potentials to differentiate the connectivity and functional architecture that distinguish neocortical areas in the adult. We conclude that the neocortical neuroepithelium generates comparable populations of cells across its extent, and when exposed to the same extrinsic cues, these populations can differentiate in comparable ways. These studies support the concept that the neocortical neuroepithelium generates a “protocortex” (20), specified to have fundamental cortical features but lacking a rigid specification of “area-specific” features. 8 1992 Academic Press, Inc.
INTRODUCTION The neocortex of adult mammals consists of numerous functionally distinct “areas” each characterized by its unique connectivity and cytoarchitecture (2, 14). Considering the striking differences that exist between areas in the adult, it is remarkable that the embryonic neocortex appears to be a uniform structure lacking area-specific distinctions. While great strides have been made in elucidating the operation and structural organization of the neocortex, only recently have researchers attended to the mechanisms by which cortical areas attain the features which distinguish one from the others (12,20,27). The query “how do cortical areas emerge?“, as with many questions in developmental biology, reduces to exploring the relative contributions of genetic
CORTICAL
TRANSPLANTATION
METHOD
As schematized in Fig. 1, a piece of late fetal rat cortex is excised and placed into a comparably sized cavity aspirated in the cortical gray matter of a neonatal host (Fig, 1). For later identification, the transplant is prelabeled with either tritiated thymidine in utero or with a fluores121 All
Copyright 0 1992 rights of reproduction
0014.4&66/92 $3.00 by Academic Press, Inc. in any form reserved.
OLFACTORY
BULB
TRANSPLANTS
129
FIG. 2. (A) Saggital Nissl-stained [3H] autoradiogram of the TX OB of Fig. 1A. Arrow indicates the coagulation lesion within the TX. B, C, and D correspond to regions illustrated in subsequent micrographs. OP is the olfactory peduncle. X18. (B) Autoradiogram of TX-lesion (L) interface. Arrows indicate examples of large neuronal nuclei with overlying silver grains, although the majority of all cell types in the TX are labeled. x310. (C) An unsuppressed silver-stained section from the same area as B. The lesion (L) is to the right and the adjacent TX area to the left has massive degenerative debris. x150. (D) Similar preparation from the OP to show a few scattered particles of degeneration (arrows) immediately under the pia (P) but otherwise normal fibers occur throughout the area. X150. (E) Silver stain from caudal OP of the same subject to show absence of degenerative material. x150. (F) Silver stain of superficial olfactory cortex ipsilateral to the TX. No degeneration but only normal fibers are seen in the molecular layer. A normally present olfactory tract is missing under the pia (P). x150.
SPECIFICATION
OF
placed embryonic motor cortex into the visual cortex of a neonatal host and examined at two survival times projections to two subcortical structures, the superior colliculus, a permanent target of visual cortex, and the spinal cord, a permanent target of sensorimotor cortex. Tracer injected into the pyramidal decussation in immature hosts retrogradely labeled layer 5 neurons in the transplanted motor cortex. However, a comparable injection into mature hosts failed to label transplanted layer 5 neurons. Importantly, though, an injection of a second retrograde tracer in the superior colliculus of the same mature hosts did label transplanted layer 5 neurons. Thus, sensorimotor layer 5 neurons, when transplanted to visual cortex, send only a transient projection to the spinal cord, but a permanent projection to the superior colliculus. In summary, heterotopically transplanted layer 5 neurons form permanent projections appropriate for their new locale. Therefore, the area-specific, selective elimination of axon collaterals is not a fixed intrinsic feature of layer 5 neurons; the remodeling of their initially widespread projections seems to depend more upon the area of cortex in which the layer 5 neurons develop than on the area in which they were generated. DEVELOPMENT WITH THALAMUS
OF AREA-SPECIFIC CONNECTIVITY AND ~ONT~I~ATE~L CORTEX
To further define the potential of cortical neurons to establish connections normally characteristic of their ~ounte~arts in other areas of cortex, we examined the thalamic and callosal connections of heterotopic transplants (19) (Fig. 3). In these experiments, a piece of late fetal visual cortex was transplanted to the sensorimotor cortex of a newborn host. When the host was an adult, WGA-HRP was iontophoresed into the transplanted cortex (Figs. 3A and 3B). In cases in which the injected tracer was confined to the transplant, retrogradely labeled cells and anterogradely labeled fibers were found in thalamic nuclei that normally form reciprocal projections with sensorimotor cortex (Fig. 3C). No labeling was present in the dorsal lateral geniculate nucleus, nor in other thalamic nuclei that project specifically to visual cortex. Additionally, labeled cells and axons were found in contralateral cortex at locations homotopic to the transplant (Fig. 3D; also see (23)). No label was present in regions of contralateral cortex homotopic to visual cortex, the source of the transplanted tissue. Thus, the thalamic and callosal connections, both afferent and efferent, formed by transplanted visual cortex are influenced by its position in the tangential plane of cortex. One explanation for the finding that the transplanted visual cortex projects exclusively to sensorimotor thalamus is that the layer 6 corticothalamic neurons had been “respecified.” A more straightforward interpretation is that layer 6 efferents from the transplanted cortex followed a pathway to the sensorimotor thalamus
NEOCORTICAL
123
AREAS
established prior to transplantation by axons from the surrounding or deleted host cortex. The most likely reason that axons from sensorimotor thalamus innervate the transplant is that they are present at that locale prior to transplantation. However, this finding shows that thalamic nuclei can innervate “foreign” pieces of cortex. Consistent with our finding is a recent co-culture study which shows that thalamocortical projections lack regional specificity in vitro (18). In normal development, appropriate thalamocortical relationships are probably promoted by cues that are either regionally specific or distributed in a graded way across the cortex; recent experimental evidence suggests that such cues are present in the cortical subplate (7). Neither the in uivo (19) nor the in u&o (18) demonstrations of the “generic” innervation of cortex by thalamus contradict this scheme of thalamocortical development. DEVELOPMENT INDICATIVE
OF AREA-SPECIFIC ARCHITE~URE OF UNIQUE FUNCTIONAL ORGANIZATIONS
The architectural ~fferentiation of a cortical area distinguishes it from other cortical areas and can reflect its functional organization. A striking example of specialized neocortical architecture is the barrelfield of the somatosensory cortex in rats and mice (38, 39). “Barrels” are discrete functional units composed of regular groupings of layer 4 neurons innervated by a cluster of thalamic afferents from the ventrobasal complex (13). Each barrel represents a single specialized sensory hair on the rodent’s body surface. The disjunctive and corresponding patterns of thalamocortical terminations and their layer 4 target cells indicative of barrels gradually emerges from uniform distributions during the first postnatal week. Two decades of experimental studies have demonstrated that normal development of the barrelfield requires an intact somatosensory periphery during a critical period of development (reviewed in (38)). However, only recently has barrel development been addressed from the perspective of whether somatosensory cortex is uniquely specified to differentiate barrels and their characteristic array (30). We examined whether a cortical region that normally never forms barrels has the capacity to develop them when transplanted to somatosensory cortex. In addition, we sought to determine whether information intrinsic to parietal cortex is necessary in the construction of the overall barrel pattern, as has been suggested ((5, 36) but see 33)). For these experiments we chose to transplant visual cortex which, like somatosensory cortex, has a granular layer 4 innervated by a substantial projection from a specific thalamic nucleus, the dorsal lateral geniculate. Visual cortex, however, has a uniform distribution of thalamocortical afferents and layer 4 neurons, in contrast to the disjunctive distribution of both of these elements in the somatosensory cortex.
124
FIG. 3. Heterotopically transplanted priate to its new position (19). (A) Darkfield labeled with tritiated thymidine (arrowheads) section adjacent to that in (A), reacted for (C) Darkfield photomicrograph of thalamus (arrow, VM) and ventrolateral (arrowhead, present in the same regions (asterisk). (D) and located in the contralateral cortex in external medullary lamina.
O’LEARY,
SCHLAGGAR,
AND
STANFIELD
visual cortex develops reciprocal connections with host thalamus and contralateral cortex approphotomicrograph of visual cortex transplant (surrounded by dashed lines), demarcated by cells exposed to tissue in hero, placed into sensorimotor cortex. (B) Brightfield photomicrograph of HRP-TMB histochemistry, to demonstrate injection of tracer limited to transplant (dashed lines). ipsilateral to injected transplant. Cells retrogradely labeled with HRP are evident in ventromedial VL) nuclei, which normally project to sensorimotor cortex. Anterogradely labeled fibers are Darkfield photomicrograph of some supragranular callosal neurons retrogradely labeled with HRP a position homotopic to the transplant. wm, white matter; iml, internal medullary lamina; eml,
Pieces of late fetal visual cortex were transplanted to the presumptive barrelfield of parietal cortex in newborn rats at ages before barrels begin to differentiate. Later, the host and transplanted cortex were assayed for barrels by reacting brain sections for acetylcholinesterase (AChE), an early marker for ventrobasal thalamic afferents (29), by peanut agglutinin (PNA) binding, a marker for glycoconjugated molecules that delineate barrel sides and septae (5), and by staining for Nissl substance, to reveal the distribution of cortical cells. Aggregates of layer 4 neurons develop in transplanted visual cortex; these cellular groupings are innervated by dense clusters of AChE-marked axons from the ventrobasal complex of the thalamus (see Fig. 2 of Ref. (30)). This experiment demonstrates that ventrobasal thalamic afferents are capable of organizing in a normal fashion in the appropriate layers of a foreign piece of cortex and that layer 4 neurons, likely in response to these afferents, redistribute into aggregates. Further, the innervation of layer 4 of transplanted visual cortex by ventrobasal afferents shows that thalamic axons have a particular affinity for layer 4 even in the absence of the
normal specific matching of thalamic nuclei with cortical areas. We conclude that the somatosensory cortex is not unique in its ability to form barrels; other cortical areas also have this potential. We also find that a normal patterning of the barrel array can develop in visual cortex transplanted to the somatosensory cortex (Fig. 4). In both host and transplanted cortex, AChE-marked barrels are ringed by PNA-labeled sides and septae. These complementary patterns are identical to those in normal somatosensory cortex. Particularly striking is the maintenance of the patterning of the barrel array across host/transplant borders. Thus not only can barrels develop in transplanted visual cortex, but their arrangement, as in normal somatosensory cortex, reflects the distribution of sensory hairs on the rat’s surface. These findings indicate that the normal, somatotopic patterning of barrels does not require cues uniquely inherent to somatosensory cortex. Further, since visual cortex has the capacity to differentiate barrels, it must not be rigidly specified to develop the architecture characteristic of visual cortex at the time it is transplanted. Similar conclusions
SPECIFICATION
OF
NEOCORTICAL
125
AREAS
FIG. 4. Development of appropriate area-specific architecture and functionai organization, i.e., patterning of barrels and glycoconjugate boundaries, in visual cortex transplanted to the barrelfield of somatosenso~ cortex. (A) In an AChE-reacted tangential section through the host somatosensory and transplanted visual cortex of a Postnatal Day 8 rat, a normal-appearing barrel pattern is evident. (B) Photomontage of PNA-stained sections adjacent to that shown in (A). The overall PNA-binding pattern resembles that in normal barrel cortex and complements the AChE-positive pattern in (A). (C). Some of the AChE-positive barrels formed in the transplanted visual cortex are numbered in this higher magnification of the section in (A). (D) In this higher magnification of the section shown in (B), glycoconjugate boundaries are identified by the same numbers that mark the corresponding AChE-positive barrels in (C). (E) Fluorescence micrograph of the same section shown in (A) taken before processing for AChE. Cells labeled with bisbenzimide mark the transplant. Dense aggregates of labeled cells are distributed in a pattern that parallels the labeling revealed by AChE histochemistry and PNA binding. The arrows in A through E indicate the same blood vessels located just outside of the transplant. Scale bar represents 0.50 mm (A and B) and 0.30 mm (C through E). Reprinted from Ref. (30) with permission. (B. L. Schlaggar and D. D. M. O’Leary. June 14,199l. Science 252:1556-1560. Copyright 1991 by the AAAS.)
can be derived from the finding that the primary visual cortex (area 17) of primates, following early bilateral enucleation, can develop a cytoarchitecture that resembles the neighboring peristriate cortex (area 18) (6,28). We conclude that somatosensory thalamocortical afferents are not only necessary for barrel formation (37), but are also sufficient to direct the differentiation of barrels and the patterning of the barrel array in normal somatosensory cortex since they can do so in transplanted visual cortex, in which no molecular framework prespecifying the development of these functional groupings can be present. CONCLUSIONS
The experimental findings considered here indicate that different regions of the developing neocortex may be interchangeable in terms of the connections they are able to develop and maintain, and even in their capacity to form complex and highly organized neuronal assemblies. We have suggested that the neocortical neuroepithelium generates across its full extent comparable populations of neurons and glia which rely on interactions later in development with extrinsic agents, for example thalamoco~ical afferents, to generate area-specific features (20,23,30,32). Our heterotopic transplant studies demonstrate that when exposed to the same extrinsic agents these populations of cells can respond in compa-
rable ways. This degree of homogeneity across the developing neocortex does not imply a complete lack of specificity. Indeed, the neocortical neuroepithelium may be highly specified to generate throughout its tangential extent repetitive arrays of the various cortical cell types (16) and a common basic circuitry (17,34), but this fundamental neocortical structure is not specified to differentiate architectural features which distinguish areas of the neocortex in adult mammals. The term “protocortex” encapsulates this model of cortical development (20), and is not merely a synonym for “undifferentiated” cortex (12). The protocortex model minimizes the need for a prespecification of the features that distinguish cortical areas (27), and provides a framework for explaining the phenomenon of cortical plasticity following trauma to the immature nervous system and the addition of new cortical areas during mammalian evolution. ACKNOWLEDGMENTS This work was supported by NISI EY07025) and NINDS (PO1 NS17763). ments on a draft of this paper.
grants from We thank
the NE1 (ROI B.B.S. for com-
REFERENCES 1.
BICKNESE, A. R., AND D. D. M. O’LEARY. potentials of layer 5 projection neurons cortex. Sot. Neurosci. A&r. 16: 311.
1990. Axon targeting across the developing
O’LEARY,
126
2. BRODMANN,
K.
ihren Principen Leipzig, FRG.
1909. Lohalisationslehre dargestellt aus Grund
SCHLAGGAR,
der Groshirnrinde in des Zellen baue Barth,
AND
22. O’LEARY,
3. CHANG, F.-L.
F., J. G. STEEDMAN, AND R. D. LUND. 1984. Emhryonic cerebral cortex placed in the occipital region of newborn rats makes connections with the host brain. Dev. Brain Res. 13: 164-166.
4. CHANG,
F.-L. F., J. G. STEEDMAN, AND R. D. LUND. 1986. The lamination and connectivity of embryonic cerebral cortex transplanted into newborn rat cortex. J. Comp. Neurol. 244: 401-411.
23.
N. G. F., AND D. A. STEINDLER. 1986. Lectins demarcate the barrel subfield in the somatosensory cortex of the early postnatal mouse. J. Comp. Neurol. 249: 157-169. DEHAY, C., G. HORSBURGH, M. BERLAND, H. KILLACKEY, AND H. KENNEDY. 1989. Maturation and connectivity of the visual cortex in monkey is altered by prenatal removal of retinal input. Nature (London) 337: 265-267.
7. GHOSH,
A., A. ANTONINI, S. K. MCCONNELL, AND C. J. SHATZ. 1990. Requirement for subplate neurons in the formation of thalamocortical connections. Nature (London) 347: 179-181.
8. HAMBURGER, Oxford
Univ.
9. INNOCENTI, establishment
V. 1988. The Heritage Press, Oxford. G. M. 1981. Growth of visual callosal
of Experimental
Embryology
and reshaping of axons in the connections. Science 212: 824-
827. 10.
Iw, G. O., AND H. P. KILLACKEY. 1982. Ontogenetic the projections of neocortical neurons. J. Neurosci.
changes in 2: 735-743.
11.
JAEGER, C. B., AND R. D. LUND. 1981. Transplantation of embryonic occipital cortex to the brain of newborn rats. An autoradiographic study of transplant histogenesis. Exp. Brain Res. 40:
265-272. 12.
KILLACKEY, ward relating 1-17.
H. P. 1990. Neocortical expansion: An attempt phylogeny and ontogeny. J. Cognitive Neurosci.
to2:
13.
KILLACKEY, H. P., AND S. LESHIN. 1975. The organization of specific thalamocortical projections to the posteromedial barrel subfield of the rat somatic sensory cortex. Brain Res. 86: 469-
25.
KRIEG, W. J. S. 1946. Connections of the cerebral cortex. I. The albino rat. A. Topography of the cortical areas. J. Comp. Neurol.
D. D. M., B. B. STANFIELD, AND W. M. COWAN. 1981. Evidence that the early postnatal restriction of the cells of origin of the callosal projection is due to the elimination of axonal collaterals rather than to the death of neurons. Dev. Brain Res. 1: 607-617. O’LEARY, D. D. M., AND T. TERASHIMA. 1988. Cortical axons branch to multiple subcortical targets by interstitial axon budding: implications for target recognition and “waiting periods.” Neuron 1: 901-910.
26. O’LEARY,
D. D. M., AND T. TERASHIMA. 1989. of cortical axons: Implications for target axons. Sot. Neurosci. Abstr. 15: 875.
branching developing
27. RAKIC, P. 1988. Specification of cerebral 241: 170-176. 28. RAKIC, P., I. SUNER, AND R. W. WILLIAMS. chitectonic tex. Proc.
30. SCHLAGGAR, visual cortex somatosensory
B. L., AND D. D. M. O’LEARY. 1991. Potential to develop an array of functional units unique cortex. Science. 252: 1556-1560.
32. STANFIELD,
B. B., AND D. D. M. O’LEARY. 1985. Fetal occipital cortical neurones transplanted to the rostra1 cortex can extend and maintain a pyramidal tract axon. Nature (London) 313: 135-137.
33. STEINDLER,
D. A., T. F. O’BRIEN, E. LAYWELL, K. HARRINGTON, A. FAISSNER, AND M. SCHACHNER. 1990. Boundaries during normal and abnormal brain development: In vivo and in vitro studies of glia and glycoconjugates. Ezp. Neurol. 109: 35-56.
34. SUR, M., P. E. GARRAGHTY,
MCCONNELL, S. K. 1988. Fates of visual ferret after isochronic and heterochronic Neurosci. 8: 945-974.
35.
17.
METIN, C., AND D. 0. FROST. 1989. Visual responses of neurons in somatosensory cortex of hamsters with experimentally induced retinal projections to somatosensory thalamus. Proc. Natl. Acad. Sci. USA 86: 357-361.
in the J.
O’LEARY, D. D. M. 1988. Visual cortex transplanted to frontal region forms reciprocal thalamic and callosal connections typical of motor cortex. Sot. Neurosci. Abstr. 14: 1113.
20. O’LEARY, 21.
D. D. M. 1989. Do cortical areas emerge from a protocortex? Trends Neurosci. 12: 400-406. O’LEARY, D. D. M., A. R. BICKNESE, J. A. DE CARLOS, C. D. HEFFNER, S. E. KOESTER, L. J. KUTKA, AND T. TERASHIMA. 1990. Target selection by cortical axons: Alternative mechanisms to establish axonal connections in the developing brain. Cold Spring Harbor Symp. Quant. Biol. Brain. 55: 453-468.
of to
B. B., D. D. M. O’LEARY, AND C. FRICKS. 1982. Selective collateral elimination in early postnatal development restricts cortical distribution of rat pyramidal tract neurons. Nature (London) 298: 371-373.
16.
19.
exsys-
31. STANFIELD,
LUND, R. D., AND M. J. MUSTARI. 1977. Development of the geniculocortical pathway in rats. J. Comp. Neurol. 173: 289-306.
MOLNAR, Z., AND C. BLAKEMORE. 1991. Lack of regional specificity for connections formed between thalamus and cortex in coculture. Nature (London) 351: 475-477.
Science
R. T. 1987. A morphogenic role for transiently pressed acetylcholinesterase in developing thalamocortical tems. Neurosci. Lett. 75: 259-264.
84: 221-215.
18.
areas.
1991. A novel cytoararea induced experimentally within the primate corNatl. Acad. Sci. USA 88: 2083-2087.
15.
cortical neurons transplantation.
cortical
Growth and selection by
29. ROBERTSON,
412. 14.
D. D. M., AND B. B. STANFIELD. 1985. Occipital cortical neurons with transient pyramidal tract axons extend and maintain collaterals to subcortical but not intracortical targets. Brain Res. 336: 326-333. O’LEARY, D. D. M., AND B. B. STANFIELD. 1989. Selective elimination of axons extended by developing cortical neurons is dependent on regional locale: experiments utilizing fetal cortical transplants. J. Neurosci. 9: 2230-2246.
24. O’LEARY,
5. COOPER,
6.
STANFIELD
AND A. W. ROE. 1988. Experimentally induced visual projections into auditory thalamus and cortex. Science 242: 1437-1441. THONG, I. G., AND B. DREHER. 1986. The development of the corticotectal pathway in the albino rat. Dev. Brain Res. 25: 227-
238. 36. WATERS,
37. 38. 39.
R. S., MCCANDLISH, C. A., AND N. G. F. COOPER. 1990. Early development of SI cortical barrel subfield representation of forelimb in normal and deafferented neonatal rat as delineated by peroxidase conjugated lectin, peanut agglutinin (PNA). Exp. Brain Res. 81: 234-240. WISE, S. P., AND E. G. JONES. 1978. Developmental studies of thalamocortical and commisural connections in the rat somatic sensory cortex. J. Comp. Neurol. 178: 187-208. WOOLSEY, T. A. 1990. Peripheral alteration and somatosensory development. In Development of Sensory Systems in Mammals (J. R. Coleman, Ed.) Wiley, New York. WOOLSEY, T. A., AND H. VAN DER Loos. 1970. The structural organization of layer IV in the somatosensory region (SI) of the mouse cerebral cortex: The description of a cortical field composed of discrete cytoarchitectonic units. Brain Res. 17: 205-
242.
115,121-126
NEUROLOGY
(1992)
The Specification of Sensory Cortex: Lessons from Cortical Transplantation DENNISD.M.O'LEARY,*BRADLEYL.SCHLAGGAR,*ANDBRENTB.STANFIELD~ *Molecular
Neurobiology Laboratory, of Mental Health,
The Sadk Institute, La Jolla, California 92037; and tNationa1 NIH Animal Center, Poolesuille, Maryland 92837
Institute
programs and environmental influences to the phenomenon of interest. In one extreme, cortical areas may be entirely prespecified-the ultimate emergence of areaspecific features is a programmed event specified in the neuroepithelium that gives rise to the cortex. In the other extreme, area-specific features may be entirely imposed upon a region of developing cortex by outside influences-the distinctive characteristics of an area result from interactions between the already assembled, yet immature cortex and agents extrinsic to the cortex. Of course, neither of these extremes is likely to be entirely correct, but they help to focus attention on an issue antecedent to an understanding of the mechanisms involved in the emergence of distinct cortical areas. Over the past several years, we have attempted to determine the extent to which the neocortical neuroepithelium is regionally specified to generate area-unique cohorts of cells programmed to develop area-specific features or, conversely, is lacking such regionalization and generates comparable populations of cells across its extent. Specifically, we have employed a heterotopic transplantation paradigm to test the potential of regions of cortex to differentiate features that are normally unique to other areas (19, 23, 30, 32). The transplantation of developing tissue, a method of great historical significance in experimental embryology and developmental neurobiology, is a powerful tool for addressing such questions of fate and potential (8). Several researchers, especially Lund and his colleagues, have demonstrated that fetal rat neocortex can be successfully transplanted: transplanted embryonic rat cortex continues to generate cells, develops a laminar structure, and forms connections with the host brain (3, 4, 11).
The mammalian neocortex is functionally organized into numerous specia!ized “areas.” The distinct functional properties characteristic of each area are in large part due to connectional and architectural differences among the areas. However, these “area-specific” features which distinguish mature areas are not apparent early in the development of neocortex. We have used heterotopic cortical transplantation to examine whether these area-specific features are prespecified or emerge as a result of epigenetic interactions. Here, we review our studies in which late fetal rat cortex was transplanted heterotopically into the cortex of newborn rats to test its capacity to differentiate features normally unique to other cortical areas. We find that regions of the developing neocortex have similar potentials to differentiate the connectivity and functional architecture that distinguish neocortical areas in the adult. We conclude that the neocortical neuroepithelium generates comparable populations of cells across its extent, and when exposed to the same extrinsic cues, these populations can differentiate in comparable ways. These studies support the concept that the neocortical neuroepithelium generates a “protocortex” (20), specified to have fundamental cortical features but lacking a rigid specification of “area-specific” features. 8 1992 Academic Press, Inc.
INTRODUCTION The neocortex of adult mammals consists of numerous functionally distinct “areas” each characterized by its unique connectivity and cytoarchitecture (2, 14). Considering the striking differences that exist between areas in the adult, it is remarkable that the embryonic neocortex appears to be a uniform structure lacking area-specific distinctions. While great strides have been made in elucidating the operation and structural organization of the neocortex, only recently have researchers attended to the mechanisms by which cortical areas attain the features which distinguish one from the others (12,20,27). The query “how do cortical areas emerge?“, as with many questions in developmental biology, reduces to exploring the relative contributions of genetic
CORTICAL
TRANSPLANTATION
METHOD
As schematized in Fig. 1, a piece of late fetal rat cortex is excised and placed into a comparably sized cavity aspirated in the cortical gray matter of a neonatal host (Fig, 1). For later identification, the transplant is prelabeled with either tritiated thymidine in utero or with a fluores121 All
Copyright 0 1992 rights of reproduction
0014.4&66/92 $3.00 by Academic Press, Inc. in any form reserved.
OLFACTORY
BULB
TRANSPLANTS
129
FIG. 2. (A) Saggital Nissl-stained [3H] autoradiogram of the TX OB of Fig. 1A. Arrow indicates the coagulation lesion within the TX. B, C, and D correspond to regions illustrated in subsequent micrographs. OP is the olfactory peduncle. X18. (B) Autoradiogram of TX-lesion (L) interface. Arrows indicate examples of large neuronal nuclei with overlying silver grains, although the majority of all cell types in the TX are labeled. x310. (C) An unsuppressed silver-stained section from the same area as B. The lesion (L) is to the right and the adjacent TX area to the left has massive degenerative debris. x150. (D) Similar preparation from the OP to show a few scattered particles of degeneration (arrows) immediately under the pia (P) but otherwise normal fibers occur throughout the area. X150. (E) Silver stain from caudal OP of the same subject to show absence of degenerative material. x150. (F) Silver stain of superficial olfactory cortex ipsilateral to the TX. No degeneration but only normal fibers are seen in the molecular layer. A normally present olfactory tract is missing under the pia (P). x150.
SPECIFICATION
OF
placed embryonic motor cortex into the visual cortex of a neonatal host and examined at two survival times projections to two subcortical structures, the superior colliculus, a permanent target of visual cortex, and the spinal cord, a permanent target of sensorimotor cortex. Tracer injected into the pyramidal decussation in immature hosts retrogradely labeled layer 5 neurons in the transplanted motor cortex. However, a comparable injection into mature hosts failed to label transplanted layer 5 neurons. Importantly, though, an injection of a second retrograde tracer in the superior colliculus of the same mature hosts did label transplanted layer 5 neurons. Thus, sensorimotor layer 5 neurons, when transplanted to visual cortex, send only a transient projection to the spinal cord, but a permanent projection to the superior colliculus. In summary, heterotopically transplanted layer 5 neurons form permanent projections appropriate for their new locale. Therefore, the area-specific, selective elimination of axon collaterals is not a fixed intrinsic feature of layer 5 neurons; the remodeling of their initially widespread projections seems to depend more upon the area of cortex in which the layer 5 neurons develop than on the area in which they were generated. DEVELOPMENT WITH THALAMUS
OF AREA-SPECIFIC CONNECTIVITY AND ~ONT~I~ATE~L CORTEX
To further define the potential of cortical neurons to establish connections normally characteristic of their ~ounte~arts in other areas of cortex, we examined the thalamic and callosal connections of heterotopic transplants (19) (Fig. 3). In these experiments, a piece of late fetal visual cortex was transplanted to the sensorimotor cortex of a newborn host. When the host was an adult, WGA-HRP was iontophoresed into the transplanted cortex (Figs. 3A and 3B). In cases in which the injected tracer was confined to the transplant, retrogradely labeled cells and anterogradely labeled fibers were found in thalamic nuclei that normally form reciprocal projections with sensorimotor cortex (Fig. 3C). No labeling was present in the dorsal lateral geniculate nucleus, nor in other thalamic nuclei that project specifically to visual cortex. Additionally, labeled cells and axons were found in contralateral cortex at locations homotopic to the transplant (Fig. 3D; also see (23)). No label was present in regions of contralateral cortex homotopic to visual cortex, the source of the transplanted tissue. Thus, the thalamic and callosal connections, both afferent and efferent, formed by transplanted visual cortex are influenced by its position in the tangential plane of cortex. One explanation for the finding that the transplanted visual cortex projects exclusively to sensorimotor thalamus is that the layer 6 corticothalamic neurons had been “respecified.” A more straightforward interpretation is that layer 6 efferents from the transplanted cortex followed a pathway to the sensorimotor thalamus
NEOCORTICAL
123
AREAS
established prior to transplantation by axons from the surrounding or deleted host cortex. The most likely reason that axons from sensorimotor thalamus innervate the transplant is that they are present at that locale prior to transplantation. However, this finding shows that thalamic nuclei can innervate “foreign” pieces of cortex. Consistent with our finding is a recent co-culture study which shows that thalamocortical projections lack regional specificity in vitro (18). In normal development, appropriate thalamocortical relationships are probably promoted by cues that are either regionally specific or distributed in a graded way across the cortex; recent experimental evidence suggests that such cues are present in the cortical subplate (7). Neither the in uivo (19) nor the in u&o (18) demonstrations of the “generic” innervation of cortex by thalamus contradict this scheme of thalamocortical development. DEVELOPMENT INDICATIVE
OF AREA-SPECIFIC ARCHITE~URE OF UNIQUE FUNCTIONAL ORGANIZATIONS
The architectural ~fferentiation of a cortical area distinguishes it from other cortical areas and can reflect its functional organization. A striking example of specialized neocortical architecture is the barrelfield of the somatosensory cortex in rats and mice (38, 39). “Barrels” are discrete functional units composed of regular groupings of layer 4 neurons innervated by a cluster of thalamic afferents from the ventrobasal complex (13). Each barrel represents a single specialized sensory hair on the rodent’s body surface. The disjunctive and corresponding patterns of thalamocortical terminations and their layer 4 target cells indicative of barrels gradually emerges from uniform distributions during the first postnatal week. Two decades of experimental studies have demonstrated that normal development of the barrelfield requires an intact somatosensory periphery during a critical period of development (reviewed in (38)). However, only recently has barrel development been addressed from the perspective of whether somatosensory cortex is uniquely specified to differentiate barrels and their characteristic array (30). We examined whether a cortical region that normally never forms barrels has the capacity to develop them when transplanted to somatosensory cortex. In addition, we sought to determine whether information intrinsic to parietal cortex is necessary in the construction of the overall barrel pattern, as has been suggested ((5, 36) but see 33)). For these experiments we chose to transplant visual cortex which, like somatosensory cortex, has a granular layer 4 innervated by a substantial projection from a specific thalamic nucleus, the dorsal lateral geniculate. Visual cortex, however, has a uniform distribution of thalamocortical afferents and layer 4 neurons, in contrast to the disjunctive distribution of both of these elements in the somatosensory cortex.
124
FIG. 3. Heterotopically transplanted priate to its new position (19). (A) Darkfield labeled with tritiated thymidine (arrowheads) section adjacent to that in (A), reacted for (C) Darkfield photomicrograph of thalamus (arrow, VM) and ventrolateral (arrowhead, present in the same regions (asterisk). (D) and located in the contralateral cortex in external medullary lamina.
O’LEARY,
SCHLAGGAR,
AND
STANFIELD
visual cortex develops reciprocal connections with host thalamus and contralateral cortex approphotomicrograph of visual cortex transplant (surrounded by dashed lines), demarcated by cells exposed to tissue in hero, placed into sensorimotor cortex. (B) Brightfield photomicrograph of HRP-TMB histochemistry, to demonstrate injection of tracer limited to transplant (dashed lines). ipsilateral to injected transplant. Cells retrogradely labeled with HRP are evident in ventromedial VL) nuclei, which normally project to sensorimotor cortex. Anterogradely labeled fibers are Darkfield photomicrograph of some supragranular callosal neurons retrogradely labeled with HRP a position homotopic to the transplant. wm, white matter; iml, internal medullary lamina; eml,
Pieces of late fetal visual cortex were transplanted to the presumptive barrelfield of parietal cortex in newborn rats at ages before barrels begin to differentiate. Later, the host and transplanted cortex were assayed for barrels by reacting brain sections for acetylcholinesterase (AChE), an early marker for ventrobasal thalamic afferents (29), by peanut agglutinin (PNA) binding, a marker for glycoconjugated molecules that delineate barrel sides and septae (5), and by staining for Nissl substance, to reveal the distribution of cortical cells. Aggregates of layer 4 neurons develop in transplanted visual cortex; these cellular groupings are innervated by dense clusters of AChE-marked axons from the ventrobasal complex of the thalamus (see Fig. 2 of Ref. (30)). This experiment demonstrates that ventrobasal thalamic afferents are capable of organizing in a normal fashion in the appropriate layers of a foreign piece of cortex and that layer 4 neurons, likely in response to these afferents, redistribute into aggregates. Further, the innervation of layer 4 of transplanted visual cortex by ventrobasal afferents shows that thalamic axons have a particular affinity for layer 4 even in the absence of the
normal specific matching of thalamic nuclei with cortical areas. We conclude that the somatosensory cortex is not unique in its ability to form barrels; other cortical areas also have this potential. We also find that a normal patterning of the barrel array can develop in visual cortex transplanted to the somatosensory cortex (Fig. 4). In both host and transplanted cortex, AChE-marked barrels are ringed by PNA-labeled sides and septae. These complementary patterns are identical to those in normal somatosensory cortex. Particularly striking is the maintenance of the patterning of the barrel array across host/transplant borders. Thus not only can barrels develop in transplanted visual cortex, but their arrangement, as in normal somatosensory cortex, reflects the distribution of sensory hairs on the rat’s surface. These findings indicate that the normal, somatotopic patterning of barrels does not require cues uniquely inherent to somatosensory cortex. Further, since visual cortex has the capacity to differentiate barrels, it must not be rigidly specified to develop the architecture characteristic of visual cortex at the time it is transplanted. Similar conclusions
SPECIFICATION
OF
NEOCORTICAL
125
AREAS
FIG. 4. Development of appropriate area-specific architecture and functionai organization, i.e., patterning of barrels and glycoconjugate boundaries, in visual cortex transplanted to the barrelfield of somatosenso~ cortex. (A) In an AChE-reacted tangential section through the host somatosensory and transplanted visual cortex of a Postnatal Day 8 rat, a normal-appearing barrel pattern is evident. (B) Photomontage of PNA-stained sections adjacent to that shown in (A). The overall PNA-binding pattern resembles that in normal barrel cortex and complements the AChE-positive pattern in (A). (C). Some of the AChE-positive barrels formed in the transplanted visual cortex are numbered in this higher magnification of the section in (A). (D) In this higher magnification of the section shown in (B), glycoconjugate boundaries are identified by the same numbers that mark the corresponding AChE-positive barrels in (C). (E) Fluorescence micrograph of the same section shown in (A) taken before processing for AChE. Cells labeled with bisbenzimide mark the transplant. Dense aggregates of labeled cells are distributed in a pattern that parallels the labeling revealed by AChE histochemistry and PNA binding. The arrows in A through E indicate the same blood vessels located just outside of the transplant. Scale bar represents 0.50 mm (A and B) and 0.30 mm (C through E). Reprinted from Ref. (30) with permission. (B. L. Schlaggar and D. D. M. O’Leary. June 14,199l. Science 252:1556-1560. Copyright 1991 by the AAAS.)
can be derived from the finding that the primary visual cortex (area 17) of primates, following early bilateral enucleation, can develop a cytoarchitecture that resembles the neighboring peristriate cortex (area 18) (6,28). We conclude that somatosensory thalamocortical afferents are not only necessary for barrel formation (37), but are also sufficient to direct the differentiation of barrels and the patterning of the barrel array in normal somatosensory cortex since they can do so in transplanted visual cortex, in which no molecular framework prespecifying the development of these functional groupings can be present. CONCLUSIONS
The experimental findings considered here indicate that different regions of the developing neocortex may be interchangeable in terms of the connections they are able to develop and maintain, and even in their capacity to form complex and highly organized neuronal assemblies. We have suggested that the neocortical neuroepithelium generates across its full extent comparable populations of neurons and glia which rely on interactions later in development with extrinsic agents, for example thalamoco~ical afferents, to generate area-specific features (20,23,30,32). Our heterotopic transplant studies demonstrate that when exposed to the same extrinsic agents these populations of cells can respond in compa-
rable ways. This degree of homogeneity across the developing neocortex does not imply a complete lack of specificity. Indeed, the neocortical neuroepithelium may be highly specified to generate throughout its tangential extent repetitive arrays of the various cortical cell types (16) and a common basic circuitry (17,34), but this fundamental neocortical structure is not specified to differentiate architectural features which distinguish areas of the neocortex in adult mammals. The term “protocortex” encapsulates this model of cortical development (20), and is not merely a synonym for “undifferentiated” cortex (12). The protocortex model minimizes the need for a prespecification of the features that distinguish cortical areas (27), and provides a framework for explaining the phenomenon of cortical plasticity following trauma to the immature nervous system and the addition of new cortical areas during mammalian evolution. ACKNOWLEDGMENTS This work was supported by NISI EY07025) and NINDS (PO1 NS17763). ments on a draft of this paper.
grants from We thank
the NE1 (ROI B.B.S. for com-
REFERENCES 1.
BICKNESE, A. R., AND D. D. M. O’LEARY. potentials of layer 5 projection neurons cortex. Sot. Neurosci. A&r. 16: 311.
1990. Axon targeting across the developing
O’LEARY,
126
2. BRODMANN,
K.
ihren Principen Leipzig, FRG.
1909. Lohalisationslehre dargestellt aus Grund
SCHLAGGAR,
der Groshirnrinde in des Zellen baue Barth,
AND
22. O’LEARY,
3. CHANG, F.-L.
F., J. G. STEEDMAN, AND R. D. LUND. 1984. Emhryonic cerebral cortex placed in the occipital region of newborn rats makes connections with the host brain. Dev. Brain Res. 13: 164-166.
4. CHANG,
F.-L. F., J. G. STEEDMAN, AND R. D. LUND. 1986. The lamination and connectivity of embryonic cerebral cortex transplanted into newborn rat cortex. J. Comp. Neurol. 244: 401-411.
23.
N. G. F., AND D. A. STEINDLER. 1986. Lectins demarcate the barrel subfield in the somatosensory cortex of the early postnatal mouse. J. Comp. Neurol. 249: 157-169. DEHAY, C., G. HORSBURGH, M. BERLAND, H. KILLACKEY, AND H. KENNEDY. 1989. Maturation and connectivity of the visual cortex in monkey is altered by prenatal removal of retinal input. Nature (London) 337: 265-267.
7. GHOSH,
A., A. ANTONINI, S. K. MCCONNELL, AND C. J. SHATZ. 1990. Requirement for subplate neurons in the formation of thalamocortical connections. Nature (London) 347: 179-181.
8. HAMBURGER, Oxford
Univ.
9. INNOCENTI, establishment
V. 1988. The Heritage Press, Oxford. G. M. 1981. Growth of visual callosal
of Experimental
Embryology
and reshaping of axons in the connections. Science 212: 824-
827. 10.
Iw, G. O., AND H. P. KILLACKEY. 1982. Ontogenetic the projections of neocortical neurons. J. Neurosci.
changes in 2: 735-743.
11.
JAEGER, C. B., AND R. D. LUND. 1981. Transplantation of embryonic occipital cortex to the brain of newborn rats. An autoradiographic study of transplant histogenesis. Exp. Brain Res. 40:
265-272. 12.
KILLACKEY, ward relating 1-17.
H. P. 1990. Neocortical expansion: An attempt phylogeny and ontogeny. J. Cognitive Neurosci.
to2:
13.
KILLACKEY, H. P., AND S. LESHIN. 1975. The organization of specific thalamocortical projections to the posteromedial barrel subfield of the rat somatic sensory cortex. Brain Res. 86: 469-
25.
KRIEG, W. J. S. 1946. Connections of the cerebral cortex. I. The albino rat. A. Topography of the cortical areas. J. Comp. Neurol.
D. D. M., B. B. STANFIELD, AND W. M. COWAN. 1981. Evidence that the early postnatal restriction of the cells of origin of the callosal projection is due to the elimination of axonal collaterals rather than to the death of neurons. Dev. Brain Res. 1: 607-617. O’LEARY, D. D. M., AND T. TERASHIMA. 1988. Cortical axons branch to multiple subcortical targets by interstitial axon budding: implications for target recognition and “waiting periods.” Neuron 1: 901-910.
26. O’LEARY,
D. D. M., AND T. TERASHIMA. 1989. of cortical axons: Implications for target axons. Sot. Neurosci. Abstr. 15: 875.
branching developing
27. RAKIC, P. 1988. Specification of cerebral 241: 170-176. 28. RAKIC, P., I. SUNER, AND R. W. WILLIAMS. chitectonic tex. Proc.
30. SCHLAGGAR, visual cortex somatosensory
B. L., AND D. D. M. O’LEARY. 1991. Potential to develop an array of functional units unique cortex. Science. 252: 1556-1560.
32. STANFIELD,
B. B., AND D. D. M. O’LEARY. 1985. Fetal occipital cortical neurones transplanted to the rostra1 cortex can extend and maintain a pyramidal tract axon. Nature (London) 313: 135-137.
33. STEINDLER,
D. A., T. F. O’BRIEN, E. LAYWELL, K. HARRINGTON, A. FAISSNER, AND M. SCHACHNER. 1990. Boundaries during normal and abnormal brain development: In vivo and in vitro studies of glia and glycoconjugates. Ezp. Neurol. 109: 35-56.
34. SUR, M., P. E. GARRAGHTY,
MCCONNELL, S. K. 1988. Fates of visual ferret after isochronic and heterochronic Neurosci. 8: 945-974.
35.
17.
METIN, C., AND D. 0. FROST. 1989. Visual responses of neurons in somatosensory cortex of hamsters with experimentally induced retinal projections to somatosensory thalamus. Proc. Natl. Acad. Sci. USA 86: 357-361.
in the J.
O’LEARY, D. D. M. 1988. Visual cortex transplanted to frontal region forms reciprocal thalamic and callosal connections typical of motor cortex. Sot. Neurosci. Abstr. 14: 1113.
20. O’LEARY, 21.
D. D. M. 1989. Do cortical areas emerge from a protocortex? Trends Neurosci. 12: 400-406. O’LEARY, D. D. M., A. R. BICKNESE, J. A. DE CARLOS, C. D. HEFFNER, S. E. KOESTER, L. J. KUTKA, AND T. TERASHIMA. 1990. Target selection by cortical axons: Alternative mechanisms to establish axonal connections in the developing brain. Cold Spring Harbor Symp. Quant. Biol. Brain. 55: 453-468.
of to
B. B., D. D. M. O’LEARY, AND C. FRICKS. 1982. Selective collateral elimination in early postnatal development restricts cortical distribution of rat pyramidal tract neurons. Nature (London) 298: 371-373.
16.
19.
exsys-
31. STANFIELD,
LUND, R. D., AND M. J. MUSTARI. 1977. Development of the geniculocortical pathway in rats. J. Comp. Neurol. 173: 289-306.
MOLNAR, Z., AND C. BLAKEMORE. 1991. Lack of regional specificity for connections formed between thalamus and cortex in coculture. Nature (London) 351: 475-477.
Science
R. T. 1987. A morphogenic role for transiently pressed acetylcholinesterase in developing thalamocortical tems. Neurosci. Lett. 75: 259-264.
84: 221-215.
18.
areas.
1991. A novel cytoararea induced experimentally within the primate corNatl. Acad. Sci. USA 88: 2083-2087.
15.
cortical neurons transplantation.
cortical
Growth and selection by
29. ROBERTSON,
412. 14.
D. D. M., AND B. B. STANFIELD. 1985. Occipital cortical neurons with transient pyramidal tract axons extend and maintain collaterals to subcortical but not intracortical targets. Brain Res. 336: 326-333. O’LEARY, D. D. M., AND B. B. STANFIELD. 1989. Selective elimination of axons extended by developing cortical neurons is dependent on regional locale: experiments utilizing fetal cortical transplants. J. Neurosci. 9: 2230-2246.
24. O’LEARY,
5. COOPER,
6.
STANFIELD
AND A. W. ROE. 1988. Experimentally induced visual projections into auditory thalamus and cortex. Science 242: 1437-1441. THONG, I. G., AND B. DREHER. 1986. The development of the corticotectal pathway in the albino rat. Dev. Brain Res. 25: 227-
238. 36. WATERS,
37. 38. 39.
R. S., MCCANDLISH, C. A., AND N. G. F. COOPER. 1990. Early development of SI cortical barrel subfield representation of forelimb in normal and deafferented neonatal rat as delineated by peroxidase conjugated lectin, peanut agglutinin (PNA). Exp. Brain Res. 81: 234-240. WISE, S. P., AND E. G. JONES. 1978. Developmental studies of thalamocortical and commisural connections in the rat somatic sensory cortex. J. Comp. Neurol. 178: 187-208. WOOLSEY, T. A. 1990. Peripheral alteration and somatosensory development. In Development of Sensory Systems in Mammals (J. R. Coleman, Ed.) Wiley, New York. WOOLSEY, T. A., AND H. VAN DER Loos. 1970. The structural organization of layer IV in the somatosensory region (SI) of the mouse cerebral cortex: The description of a cortical field composed of discrete cytoarchitectonic units. Brain Res. 17: 205-
242.
