- Email: [email protected]
Visual development: Making maps in the dark
R342
Dispatch
Visual development: Making maps in the dark Mark Hübener
that the receptive-field properties of such neurons are modifiable by visual experience [1]. Subsequently, a large number of studies have shown that the response properties of neurons in the visual cortex are, to some extent, shaped by vision during a limited phase after birth known as the ‘critical period’. As neuronal response properties vary in an orderly fashion across the cortical surface, forming what are known as ‘cortical maps’, a change in the response characteristics of individual cells is also expected to alter the layout of these maps to a certain degree (Figure 1).
Both intrinsic and visually evoked neural activity have been suggested to play a part in development of the mammalian visual system; recent results suggest that intrinsic activity is more important than previously thought, but much remains to be learned about the contributions of other factors. Address: Max-Planck-Institut für Neurobiologie, Am Klopferspitz 18a, 82152 Martinsried, Germany. E-mail: [email protected] Current Biology 1998, 8:R342–R345 http://biomednet.com/elecref/09609822008R0342
Although the precise mechanisms underlying these modifications are still elusive, there is good evidence that competitive synaptic interactions play a crucial role. For instance, if visual input is restricted to one eye — ‘monocular deprivation’ — during the critical period,
© Current Biology Ltd ISSN 0960-9822
Shortly after their first description of receptive fields of neurons in the visual cortex, Hubel and Wiesel discovered Figure 1
Orientation preference
Ocular dominance Strabismic
MD
Normal
Stripe-reared
Number of cells
Normal
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
Current Biology
Visual experience influences the response properties of cortical neurons. The cartoons (top) indicate the type of experimental manipulation carried out during the ‘critical period’ of visual development. The histograms (middle) show the proportions of cells in different response classes, as assessed with electrical recordings. Ocular dominance: classes 1 and 5, strong contralateral or ipsilateral eye dominance, respectively; class 3, equally well driven by either eye; classes 2 and 4, intermediate. For orientation preference, the preferred orientation is indicated by the colored bar. At the bottom, the changes in map layout observed after manipulating the visual input are shown schematically. In normally reared kittens, the ocular dominance distribution is roughly symmetric, and approximately equal areas of the cortex are innervated by afferents from each eye, leading to the familiar
pattern of ocular dominance columns. In monocularly deprived (MD) animals, most cells respond strongly to the open eye, with a corresponding enlargement of this eye’s columns. In cats with a surgically induced misalignment of the optical axes of the two eyes (‘strabismic’ cats), the proportion of binocular neurons drops dramatically and the ocular dominance columns are more distinct, with sharper borders. Each orientation has an approximately equal representation in the visual cortex of normal cats, with the orientation domains arranged in a pinwheel-like fashion. Exposing kittens to only one orientation leads to a relative increase in the proportion of neurons preferring this orientation. Domains activated by this orientation are enlarged, but the basic layout of the orientation preference map is not changed [22].
Dispatch
cortical connections from that eye weaken relative to those from the other eye, but continuous pharmacological blockade of N-methyl-D-aspartate (NMDA) receptors during this period rescues the input from the closed eye [2]. Indeed, a ‘paradoxical’ shift was found to occur in favor of the deprived eye after complete pharmacological silencing of cortical neurons [3].
R343
to occur even after prenatal ocular enucleation performed at a time before photoreceptors had been generated [7].
These observations suggest that, during the critical period, inputs from the two eyes compete for influence on cortical cells, the competition being mediated by changes in synaptic efficacy. These changes have been suggested to occur by a ‘Hebbian’ mechanism, in which synaptic efficacy is modified according to the relative timing of activity in the cortical cell and in the presynaptic fibers carrying input from either eye. Probably the most direct evidence for a mechanism in which a synapse is strengthened if there is a positive correlation between activity in the presynaptic and postsynaptic cells comes from experiments where the response of a cortical cell to a specific visual stimulus was enhanced for hours after repeatedly pairing this stimulus with depolarization of the cell [4].
Recent experiments using optical imaging have now shown that the orientation map in the visual cortex — the systematic variation in orientation preference over the cortical surface — also develops without patterned visual experience. Crair et al. [8] deprived cats binocularly before the time of eye opening and found that they developed orderly orientation preference maps that initially were similar to the ones imaged in normal animals (Figure 2). Only after prolonged binocular deprivation did the maps deteriorate. These results are similar to those obtained earlier by Gödecke et al. [9], and related results were found by Chapman et al. [10] in the ferret, where at least one animal showed a faint orientation map before natural eye opening. These two studies also showed that, after the first appearance of an orientation map, its basic layout is very stable during normal development [9,10]. Taken together, these observations provide clear evidence that the initial formation of maps in the visual cortex does not depend on sensory experience.
The multitude of experiments demonstrating an influence of the visual environment on functional properties of cortical neurons has led to the common notion that visual experience is also instrumental for setting up the cortical maps. Early studies, however, had already shown that key receptive-field properties, such as orientation selectivity, are present in at least a fraction of cortical cells before the onset of vision. Moreover, two important features of the cortical functional architecture, ‘blobs’ — regularly spaced patches of relatively high cytochrome oxidase activity — and ocular dominance columns — alternating regions predominantly excited by one or other eye — were observed in primates at or before birth [5,6] (Figure 2). And the formation of cytochrome oxidase blobs was shown
How does the presence of elaborated maps in visually naïve animals fit with the observations demonstrating that vision can to some extent influence the structure of cortical maps? Are the mechanisms that generate these early maps entirely different from those responsible for altering the maps in response to visual input? A new view on these questions has come from the observation that nerve cells in the visual system are spontaneously active during early development. While the presence of this neural activity had been described before [11], it has become clear only recently that this activity shares important features with visually evoked action potentials, namely that it is patterned in space and time. It was shown that, in early postnatal ferrets, waves of excitation sweep across the retina
Figure 2 (a)
(b)
Maps in the visual cortex develop without visual experience. (a) Ocular dominance columns and (b) cytochrome oxidase blobs in the visual cortex of a newborn macaque monkey deprived of visual experience. (Reproduced with permission from [6].) (c) Orientation preference
(c)
map in the visual cortex of a kitten that was binocularly deprived before natural eye opening. The preferred orientation is color-coded according to the scheme on the right of the figure. (Image provided by Michael Crair.) Scale bars, 5 mm in (a) and (b), 1 mm in (c).
R344
Current Biology, Vol 8 No 10
long before the eyes open [12]. Now Penn et al. [13] have obtained evidence that the retinal waves are instrumental in setting up the correct wiring in another part of the visual system, the lateral geniculate nucleus (LGN) that acts as a ‘relay station’ between the eyes and visual cortex. In the LGN, the segregation of retinal ganglion cell axons into eye-specific layers is sculpted from an initially intermingled termination pattern by a process that has previously been shown to be activity-dependent [14]. The important observation made by Penn et al. [13] is that blocking the waves in one eye results in an enlargement of the projection zone of the active eye, while the area occupied by fibers from the inactive eye is strongly reduced. Thus, besides confirming that the segregation in the LGN is activity-dependent, these experiments also clearly demonstrate that this process involves competition between fibers from the two eyes. Does spontaneous activity in the visual system also contribute to the generation of cortical orientation selectivity and the formation of the orientation map? Intuitively, the retinal waves seem to be ideally suited for such a role, as they could mimic the oriented, moving contours that are later so effective in driving cortical cells. Weliky and Katz [15], in a technically challenging experiment, tackled this question in the developing ferret visual system by artificially generating correlated activity in many ganglion cells, through chronic optic nerve stimulation. They found that orientation tuning of most cortical neurons was considerably weakened, indicating that neural activity indeed plays an instructive role in the development of orientation selectivity. Somewhat surprisingly, however, it turned out that the layout of the orientation maps in the stimulated animals was fairly normal. One might argue that the artificial stimulation in these experiments started too late — towards the end of the period of traveling retinal waves [16] — to effectively alter the development of the orientation map. But the view that retinal waves are instructing the layout of the orientation map, as appealing as it might be, is difficult to reconcile with yet another observation, namely that both eyes develop identical maps in the cortex of binocularly deprived animals [8,9]. This match would require either that the waves travel in synchronized fashion in both eyes — which seems extremely unlikely and is incompatible with the role the waves are thought to play in the eyespecific segregation in the LGN — or that patterned activity in one eye is conveyed to the cortex via LGN neurons while they are still binocularly driven, thereby enabling both eyes to establish identical orientation maps in the cortex. The latter scenario, however, would predict that an initial orientation map is set up in the cortex before the LGN layers are fully segregated at around postnatal day 21 in the ferret. At present, there are no indications that an orientation map exists in the visual cortex this early
[10]. Thus, while the retinal waves seem to play a role in the maturation of the tuning properties of cortical cells, they might be less important when it comes to the more global organization of the cortical orientation map. As it appears difficult to account for all the properties of orientation maps by activity-dependent processes, one might conclude that other mechanisms are involved. In the retinotectal system, there is good evidence that the formation of another map — the retinotopic map in which neighbouring parts of the retina are represented by neighbouring parts of the optic tectum — largely relies on molecular cues [17]. Could molecules set up the orientation map? Again, this seems unlikely. Whereas the retinotopic map is based on a comparatively simple point-to-point projection, orientation selectivity is thought to arise from the convergence of LGN fibers having receptive fields that are aligned in the visual field [18]. Consequently, each LGN neuron can connect to cortical cells with different orientation preferences, depending on the relative position of its receptive field in visual space. It is hard to imagine how such intricate connections could be brought about by a molecular code. Furthermore, no markers have yet been found in the visual cortex with a distribution that correlates in any way with the system of orientation domains. There is, however, yet another potential player that should be considered: a system of clustered horizontal connections links domains with similar orientation preference in the adult visual cortex. During development, these horizontal connections first form crude clusters, a few days before orientation maps can be detected [19,20]. Later, the clusters are refined by a process that requires patterned visual input. Interestingly, the formation of the crude clusters is not prevented by the complete blockade of retinal activity by tetrodotoxin injections or ocular enucleation (though intracortical tetrodotoxin infusion prevents their emergence) [20,21]. Thus, without any activity coming from the retina, the cortex is capable of establishing a set of connections that could act as a scaffold for the developing orientation map. The spacing and arrangement of these clusters would thus determine the basic layout of the orientation map, without specifying the exact orientation preference for every cortical locus. Unfortunately, without making further assumptions this scheme does not explain how both eyes come to develop similar orientation maps. It turns out, then, that we are far from understanding to what extent activity-dependent mechanisms contribute to the development of orientation maps in particular, and cortical modules in general. One possible approach to this question currently taken is the study of maps in genetically identical individuals (T. Bonhoeffer, personal communication). While it is safe to predict that the layout of cortical maps is under the influence of both intrinsic and
Dispatch
activity-dependent factors, such experiments might teach us what their relative contributions are. Acknowledgements I thank Frank Brinkmann for help with the figures, and Tobias Bonhoeffer and Frank Sengpiel for valuable comments on the manuscript. I also thank Jonathan Horton and Michael Crair for providing figures.
References 1. Wiesel TN, Hubel DH: Single-cell responses in striate cortex of kittens deprived of vision in one eye. J Neurophysiol 1963, 26:1003-1017. 2. Kleinschmidt A, Bear MF, Singer W: Blockade of “NMDA” receptors disrupts experience-dependent plasticity of kitten striate cortex. Science 1987, 238:355-358. 3. Reiter HO, Stryker MP: Neural plasticity without postsynaptic action potentials: less-active inputs become dominant when kitten visual cortical cells are pharmacologically inhibited. Proc Natl Acad Sci USA 1988, 85:3623-3627. 4. Frégnac Y, Schulz D, Thorpe S, Bienenstock E: A cellular analogue of visual cortical plasticity. Nature 1988, 333:367-370. 5. Horton JC: Cytochrome oxidase patches: a new cytoarchitectonic feature of monkey visual cortex. Phil Trans R Soc Lond [Biol] 1984, 304:199-253. 6. Horton JC, Hocking DR: An adult-like pattern of ocular dominance columns in striate cortex of newborn monkeys prior to visual experience. J Neurosci 1996, 16:1791-1807. 7. Kuljis RO, Rakic P: Hypercolumns in primate visual cortex can develop in the absence of cues from photoreceptors. Proc Natl Acad Sci USA 1990, 87:5303-5306. 8. Crair MC, Gillespie DC, Stryker MP: The role of visual experience in the development of columns in cat visual cortex. Science 1998, 279:566-570. 9. Gödecke I, Kim DS, Bonhoeffer T, Singer W: Development of orientation preference maps in area 18 of kitten visual cortex. Eur J Neurosci 1997, 9:1754-1762. 10. Chapman B, Stryker MP, Bonhoeffer T: Development of orientation preference maps in ferret primary visual cortex. J Neurosci 1996, 16:6443-6453. 11. Galli L, Maffei L: Spontaneous impulse activity of rat retinal ganglion cells in prenatal life. Science 1988, 242:90-91. 12. Meister M, Wong ROL, Baylor DA, Shatz CJ: Synchronous bursts of action potentials in ganglion cells of the developing mammalian retina. Science 1991, 252:939-943. 13. Penn AA, Riquelme PA, Feller MB, Shatz CJ: Competition in retinogeniculate patterning driven by spontaneous activity. Science 1998, 279:2108-2112. 14. Shatz CJ, Stryker MP: Prenatal tetrodotoxin infusion blocks segregation of retinogeniculate afferents. Science 1988, 242:87-89. 15. Weliky M, Katz LC: Disruption of orientation tuning in visual cortex by artificially correlated neuronal activity. Nature 1997, 386:680-685. 16. Wong ROL, Meister M, Shatz CJ: Transient period of correlated bursting activity during development of the mammalian retina. Neuron 1993, 11:923-938. 17. Drescher U, Bonhoeffer F, Müller BK: The Eph family in retinal axon guidance. Curr Opin Neurobiol 1997, 7:75-80. 18. Hubel DH, Wiesel TN: Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. J Physiol 1962, 160:106-154. 19. Callaway EM, Katz LC: Emergence and refinement of clustered horizontal connections in cat striate cortex. J Neurosci 1990, 10:1134-1153. 20. Ruthazer ES, Stryker MP: The role of activity in the development of long-range horizontal connections in area 17 of the ferret. J Neurosci 1996, 16:7253-7269. 21. Callaway EM, Katz LC: Effects of binocular deprivation on the development of clustered horizontal connections in cat striate cortex. Proc Natl Acad Sci USA 1991, 88:745-749. 22. Stawinski P, Sengpiel F, Bonhoeffer T: Optical imaging reveals effects of a single-orientation environment on orientation maps in cat visual cortex. Soc Neurosci Abstr 1997, 23:1027.
R345
Dispatch
Visual development: Making maps in the dark Mark Hübener
that the receptive-field properties of such neurons are modifiable by visual experience [1]. Subsequently, a large number of studies have shown that the response properties of neurons in the visual cortex are, to some extent, shaped by vision during a limited phase after birth known as the ‘critical period’. As neuronal response properties vary in an orderly fashion across the cortical surface, forming what are known as ‘cortical maps’, a change in the response characteristics of individual cells is also expected to alter the layout of these maps to a certain degree (Figure 1).
Both intrinsic and visually evoked neural activity have been suggested to play a part in development of the mammalian visual system; recent results suggest that intrinsic activity is more important than previously thought, but much remains to be learned about the contributions of other factors. Address: Max-Planck-Institut für Neurobiologie, Am Klopferspitz 18a, 82152 Martinsried, Germany. E-mail: [email protected] Current Biology 1998, 8:R342–R345 http://biomednet.com/elecref/09609822008R0342
Although the precise mechanisms underlying these modifications are still elusive, there is good evidence that competitive synaptic interactions play a crucial role. For instance, if visual input is restricted to one eye — ‘monocular deprivation’ — during the critical period,
© Current Biology Ltd ISSN 0960-9822
Shortly after their first description of receptive fields of neurons in the visual cortex, Hubel and Wiesel discovered Figure 1
Orientation preference
Ocular dominance Strabismic
MD
Normal
Stripe-reared
Number of cells
Normal
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
Current Biology
Visual experience influences the response properties of cortical neurons. The cartoons (top) indicate the type of experimental manipulation carried out during the ‘critical period’ of visual development. The histograms (middle) show the proportions of cells in different response classes, as assessed with electrical recordings. Ocular dominance: classes 1 and 5, strong contralateral or ipsilateral eye dominance, respectively; class 3, equally well driven by either eye; classes 2 and 4, intermediate. For orientation preference, the preferred orientation is indicated by the colored bar. At the bottom, the changes in map layout observed after manipulating the visual input are shown schematically. In normally reared kittens, the ocular dominance distribution is roughly symmetric, and approximately equal areas of the cortex are innervated by afferents from each eye, leading to the familiar
pattern of ocular dominance columns. In monocularly deprived (MD) animals, most cells respond strongly to the open eye, with a corresponding enlargement of this eye’s columns. In cats with a surgically induced misalignment of the optical axes of the two eyes (‘strabismic’ cats), the proportion of binocular neurons drops dramatically and the ocular dominance columns are more distinct, with sharper borders. Each orientation has an approximately equal representation in the visual cortex of normal cats, with the orientation domains arranged in a pinwheel-like fashion. Exposing kittens to only one orientation leads to a relative increase in the proportion of neurons preferring this orientation. Domains activated by this orientation are enlarged, but the basic layout of the orientation preference map is not changed [22].
Dispatch
cortical connections from that eye weaken relative to those from the other eye, but continuous pharmacological blockade of N-methyl-D-aspartate (NMDA) receptors during this period rescues the input from the closed eye [2]. Indeed, a ‘paradoxical’ shift was found to occur in favor of the deprived eye after complete pharmacological silencing of cortical neurons [3].
R343
to occur even after prenatal ocular enucleation performed at a time before photoreceptors had been generated [7].
These observations suggest that, during the critical period, inputs from the two eyes compete for influence on cortical cells, the competition being mediated by changes in synaptic efficacy. These changes have been suggested to occur by a ‘Hebbian’ mechanism, in which synaptic efficacy is modified according to the relative timing of activity in the cortical cell and in the presynaptic fibers carrying input from either eye. Probably the most direct evidence for a mechanism in which a synapse is strengthened if there is a positive correlation between activity in the presynaptic and postsynaptic cells comes from experiments where the response of a cortical cell to a specific visual stimulus was enhanced for hours after repeatedly pairing this stimulus with depolarization of the cell [4].
Recent experiments using optical imaging have now shown that the orientation map in the visual cortex — the systematic variation in orientation preference over the cortical surface — also develops without patterned visual experience. Crair et al. [8] deprived cats binocularly before the time of eye opening and found that they developed orderly orientation preference maps that initially were similar to the ones imaged in normal animals (Figure 2). Only after prolonged binocular deprivation did the maps deteriorate. These results are similar to those obtained earlier by Gödecke et al. [9], and related results were found by Chapman et al. [10] in the ferret, where at least one animal showed a faint orientation map before natural eye opening. These two studies also showed that, after the first appearance of an orientation map, its basic layout is very stable during normal development [9,10]. Taken together, these observations provide clear evidence that the initial formation of maps in the visual cortex does not depend on sensory experience.
The multitude of experiments demonstrating an influence of the visual environment on functional properties of cortical neurons has led to the common notion that visual experience is also instrumental for setting up the cortical maps. Early studies, however, had already shown that key receptive-field properties, such as orientation selectivity, are present in at least a fraction of cortical cells before the onset of vision. Moreover, two important features of the cortical functional architecture, ‘blobs’ — regularly spaced patches of relatively high cytochrome oxidase activity — and ocular dominance columns — alternating regions predominantly excited by one or other eye — were observed in primates at or before birth [5,6] (Figure 2). And the formation of cytochrome oxidase blobs was shown
How does the presence of elaborated maps in visually naïve animals fit with the observations demonstrating that vision can to some extent influence the structure of cortical maps? Are the mechanisms that generate these early maps entirely different from those responsible for altering the maps in response to visual input? A new view on these questions has come from the observation that nerve cells in the visual system are spontaneously active during early development. While the presence of this neural activity had been described before [11], it has become clear only recently that this activity shares important features with visually evoked action potentials, namely that it is patterned in space and time. It was shown that, in early postnatal ferrets, waves of excitation sweep across the retina
Figure 2 (a)
(b)
Maps in the visual cortex develop without visual experience. (a) Ocular dominance columns and (b) cytochrome oxidase blobs in the visual cortex of a newborn macaque monkey deprived of visual experience. (Reproduced with permission from [6].) (c) Orientation preference
(c)
map in the visual cortex of a kitten that was binocularly deprived before natural eye opening. The preferred orientation is color-coded according to the scheme on the right of the figure. (Image provided by Michael Crair.) Scale bars, 5 mm in (a) and (b), 1 mm in (c).
R344
Current Biology, Vol 8 No 10
long before the eyes open [12]. Now Penn et al. [13] have obtained evidence that the retinal waves are instrumental in setting up the correct wiring in another part of the visual system, the lateral geniculate nucleus (LGN) that acts as a ‘relay station’ between the eyes and visual cortex. In the LGN, the segregation of retinal ganglion cell axons into eye-specific layers is sculpted from an initially intermingled termination pattern by a process that has previously been shown to be activity-dependent [14]. The important observation made by Penn et al. [13] is that blocking the waves in one eye results in an enlargement of the projection zone of the active eye, while the area occupied by fibers from the inactive eye is strongly reduced. Thus, besides confirming that the segregation in the LGN is activity-dependent, these experiments also clearly demonstrate that this process involves competition between fibers from the two eyes. Does spontaneous activity in the visual system also contribute to the generation of cortical orientation selectivity and the formation of the orientation map? Intuitively, the retinal waves seem to be ideally suited for such a role, as they could mimic the oriented, moving contours that are later so effective in driving cortical cells. Weliky and Katz [15], in a technically challenging experiment, tackled this question in the developing ferret visual system by artificially generating correlated activity in many ganglion cells, through chronic optic nerve stimulation. They found that orientation tuning of most cortical neurons was considerably weakened, indicating that neural activity indeed plays an instructive role in the development of orientation selectivity. Somewhat surprisingly, however, it turned out that the layout of the orientation maps in the stimulated animals was fairly normal. One might argue that the artificial stimulation in these experiments started too late — towards the end of the period of traveling retinal waves [16] — to effectively alter the development of the orientation map. But the view that retinal waves are instructing the layout of the orientation map, as appealing as it might be, is difficult to reconcile with yet another observation, namely that both eyes develop identical maps in the cortex of binocularly deprived animals [8,9]. This match would require either that the waves travel in synchronized fashion in both eyes — which seems extremely unlikely and is incompatible with the role the waves are thought to play in the eyespecific segregation in the LGN — or that patterned activity in one eye is conveyed to the cortex via LGN neurons while they are still binocularly driven, thereby enabling both eyes to establish identical orientation maps in the cortex. The latter scenario, however, would predict that an initial orientation map is set up in the cortex before the LGN layers are fully segregated at around postnatal day 21 in the ferret. At present, there are no indications that an orientation map exists in the visual cortex this early
[10]. Thus, while the retinal waves seem to play a role in the maturation of the tuning properties of cortical cells, they might be less important when it comes to the more global organization of the cortical orientation map. As it appears difficult to account for all the properties of orientation maps by activity-dependent processes, one might conclude that other mechanisms are involved. In the retinotectal system, there is good evidence that the formation of another map — the retinotopic map in which neighbouring parts of the retina are represented by neighbouring parts of the optic tectum — largely relies on molecular cues [17]. Could molecules set up the orientation map? Again, this seems unlikely. Whereas the retinotopic map is based on a comparatively simple point-to-point projection, orientation selectivity is thought to arise from the convergence of LGN fibers having receptive fields that are aligned in the visual field [18]. Consequently, each LGN neuron can connect to cortical cells with different orientation preferences, depending on the relative position of its receptive field in visual space. It is hard to imagine how such intricate connections could be brought about by a molecular code. Furthermore, no markers have yet been found in the visual cortex with a distribution that correlates in any way with the system of orientation domains. There is, however, yet another potential player that should be considered: a system of clustered horizontal connections links domains with similar orientation preference in the adult visual cortex. During development, these horizontal connections first form crude clusters, a few days before orientation maps can be detected [19,20]. Later, the clusters are refined by a process that requires patterned visual input. Interestingly, the formation of the crude clusters is not prevented by the complete blockade of retinal activity by tetrodotoxin injections or ocular enucleation (though intracortical tetrodotoxin infusion prevents their emergence) [20,21]. Thus, without any activity coming from the retina, the cortex is capable of establishing a set of connections that could act as a scaffold for the developing orientation map. The spacing and arrangement of these clusters would thus determine the basic layout of the orientation map, without specifying the exact orientation preference for every cortical locus. Unfortunately, without making further assumptions this scheme does not explain how both eyes come to develop similar orientation maps. It turns out, then, that we are far from understanding to what extent activity-dependent mechanisms contribute to the development of orientation maps in particular, and cortical modules in general. One possible approach to this question currently taken is the study of maps in genetically identical individuals (T. Bonhoeffer, personal communication). While it is safe to predict that the layout of cortical maps is under the influence of both intrinsic and
Dispatch
activity-dependent factors, such experiments might teach us what their relative contributions are. Acknowledgements I thank Frank Brinkmann for help with the figures, and Tobias Bonhoeffer and Frank Sengpiel for valuable comments on the manuscript. I also thank Jonathan Horton and Michael Crair for providing figures.
References 1. Wiesel TN, Hubel DH: Single-cell responses in striate cortex of kittens deprived of vision in one eye. J Neurophysiol 1963, 26:1003-1017. 2. Kleinschmidt A, Bear MF, Singer W: Blockade of “NMDA” receptors disrupts experience-dependent plasticity of kitten striate cortex. Science 1987, 238:355-358. 3. Reiter HO, Stryker MP: Neural plasticity without postsynaptic action potentials: less-active inputs become dominant when kitten visual cortical cells are pharmacologically inhibited. Proc Natl Acad Sci USA 1988, 85:3623-3627. 4. Frégnac Y, Schulz D, Thorpe S, Bienenstock E: A cellular analogue of visual cortical plasticity. Nature 1988, 333:367-370. 5. Horton JC: Cytochrome oxidase patches: a new cytoarchitectonic feature of monkey visual cortex. Phil Trans R Soc Lond [Biol] 1984, 304:199-253. 6. Horton JC, Hocking DR: An adult-like pattern of ocular dominance columns in striate cortex of newborn monkeys prior to visual experience. J Neurosci 1996, 16:1791-1807. 7. Kuljis RO, Rakic P: Hypercolumns in primate visual cortex can develop in the absence of cues from photoreceptors. Proc Natl Acad Sci USA 1990, 87:5303-5306. 8. Crair MC, Gillespie DC, Stryker MP: The role of visual experience in the development of columns in cat visual cortex. Science 1998, 279:566-570. 9. Gödecke I, Kim DS, Bonhoeffer T, Singer W: Development of orientation preference maps in area 18 of kitten visual cortex. Eur J Neurosci 1997, 9:1754-1762. 10. Chapman B, Stryker MP, Bonhoeffer T: Development of orientation preference maps in ferret primary visual cortex. J Neurosci 1996, 16:6443-6453. 11. Galli L, Maffei L: Spontaneous impulse activity of rat retinal ganglion cells in prenatal life. Science 1988, 242:90-91. 12. Meister M, Wong ROL, Baylor DA, Shatz CJ: Synchronous bursts of action potentials in ganglion cells of the developing mammalian retina. Science 1991, 252:939-943. 13. Penn AA, Riquelme PA, Feller MB, Shatz CJ: Competition in retinogeniculate patterning driven by spontaneous activity. Science 1998, 279:2108-2112. 14. Shatz CJ, Stryker MP: Prenatal tetrodotoxin infusion blocks segregation of retinogeniculate afferents. Science 1988, 242:87-89. 15. Weliky M, Katz LC: Disruption of orientation tuning in visual cortex by artificially correlated neuronal activity. Nature 1997, 386:680-685. 16. Wong ROL, Meister M, Shatz CJ: Transient period of correlated bursting activity during development of the mammalian retina. Neuron 1993, 11:923-938. 17. Drescher U, Bonhoeffer F, Müller BK: The Eph family in retinal axon guidance. Curr Opin Neurobiol 1997, 7:75-80. 18. Hubel DH, Wiesel TN: Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. J Physiol 1962, 160:106-154. 19. Callaway EM, Katz LC: Emergence and refinement of clustered horizontal connections in cat striate cortex. J Neurosci 1990, 10:1134-1153. 20. Ruthazer ES, Stryker MP: The role of activity in the development of long-range horizontal connections in area 17 of the ferret. J Neurosci 1996, 16:7253-7269. 21. Callaway EM, Katz LC: Effects of binocular deprivation on the development of clustered horizontal connections in cat striate cortex. Proc Natl Acad Sci USA 1991, 88:745-749. 22. Stawinski P, Sengpiel F, Bonhoeffer T: Optical imaging reveals effects of a single-orientation environment on orientation maps in cat visual cortex. Soc Neurosci Abstr 1997, 23:1027.
R345