This essay looks at six APS classic papers published by D. H. Hubel and T. N. Wiesel that first identified a developmental critical period for environment influenced receptive field plasticity in the visual pathway. These classic papers are freely available online. These are listed here, in chronological order.
Wiesel TN, Hubel DH. Effects of visual deprivation on morphology and physiology of cells in the cat's lateral geniculate body. J Neurophysiol 26: 978–993, 1963 (http://jn.physiology.org/cgi/reprint/26/6/978).
Hubel DH, Wiesel TN. Receptive fields of cells in striate cortex of very young, visually inexperienced kittens. J Neurophysiol 26: 994–1002, 1963 (http://jn.physiology.org/cgi/reprint/26/6/994).
Wiesel TN, Hubel DH. Single-cell responses in striate cortex of kittens deprived of vision in one eye. J Neurophysiol 26: 1003–1017, 1963 (http://jn.physiology.org/cgi/reprint/26/6/1003).
Wiesel TN, Hubel DH. Comparison of the effects of unilateral and bilateral eye closure on cortical unit responses in kittens. J Neurophysiol 28: 1029–1040, 1965 (http://jn.physiology.org/cgi/reprint/28/6/1029).
Hubel DH, Wiesel TN. Binocular interaction in striate cortex of kittens reared with artificial squint. J Neurophysiol 28: 1041–1059, 1965 (http://jn.physiology.org/cgi/reprint/28/6/1041).
Wiesel TN, Hubel DH. Extent of recovery from the effects of visual deprivation in kittens. J Neurophysiol 28: 1060–1072, 1965 (http://jn.physiology.org/cgi/reprint/28/6/1060).
the following sentence opens Torsten Wiesel's 1981 Nobel Lecture: “In the early sixties, having begun to describe the physiology of cells in the adult cat visual cortex, David Hubel and I decided to investigate how the highly specific response properties of cortical cells emerged during postnatal development” (http://nobelprize.org/nobel_prizes/medicine/laureates/1981/wiesel-lecture.html). This modest statement belies the tidal wave of experiments on developmental brain plasticity that was initiated by their publications in 1963, continues to this day, and extends well beyond the occipital lobe to virtually all sensory areas, motor areas, and to “higher centers” involved in learning, memory, and decision making. To be sure, many great experimentalists and philosophers dealt with the issue of nature versus nurture in brain development before them (see Lehrman 1970), but Wiesel and Hubel were the first to recognize that their high-resolution, single-neuron analyses of visual response properties in cat cortex (Hubel and Wiesel 1962, 1963a) provided a powerful approach to deciphering how much of sensory feature extraction was fixed at birth and how much depended for its appearance on environment-introduced visual activity.
The six papers published in the Journal of Neurophysiology between 1963 and 1965 constitute the seminal contributions of David Hubel and Torsten Wiesel to developmental plasticity in the central nervous system. The work grew out of Stephen Kuffler's successful effort to bring the two together to work on the central visual pathway of the cat in his laboratory in the Wilmer Institute of Ophthalmology at the Johns Hopkins Hospital. Kuffler had already distinguished himself in visual neurophysiology by characterizing the “on” and “off” center-surround receptive field properties of ganglion cells in the cat retina (Kuffler 1953). In Kuffler's laboratory, quite by accident, according to David Hubel (http://nobelprize.org/nobel_prizes/medicine/laureates/1981/hubel-lecture.html), he and Torsten discovered that cat striate cortical neurons responded dramatically to the oriented edge of a clear microscope slide, ignoring the small black dot they had pasted on the slide hoping that it would drive the cortical neurons as effectively as dots drove the ganglion cells in Kuffler's original study. Moreover, not any edge would do. The edge had to be presented at a particular position in the visual field (expected) and at an angle specific to that cell (unexpected). The paper describing the visual cortical simple cells (Hubel and Wiesel 1959) was the beginning of a series of publications that described cortical neuron responsiveness in terms of binocularity, edge, bar, and slit detection, and orientation selectivity (Hubel and Wiesel, 1962, 1963a). From this work and subsequent combined anatomical and physiological studies on cats and monkeys (Hubel and Wiesel, 1977; LeVay et al., 1975; Livingstone and Hubel, 1984) the notion of a cortical hypercolumn would eventually arise as a representation of one locus in visual space encompassing cells with color sensitivity, all orientation sensitivities, and receiving input of varying degrees through both eyes. In the interim, and shortly after a move to Harvard University, Hubel and Wiesel began their developmental work.
Studies in visually naive kittens
In the first paper of the 1963 triple back-to-back series (Wiesel and Hubel 1963a), one eyelid of kittens was sutured closed just before eye opening. The kittens were reared in this monocularly deprived condition until they were 3 months old, at which time electrophysiological analyses of afferent retinal axons and neurons in the lateral geniculate nucleus (LGN) contralateral to the deprived eye were undertaken. In general, the visual receptive field properties of the LGN neurons were similar to those of adult cats in size and center-surround organization, although, in the visually deprived layers, background activity was low and few single-cell responses could be isolated. Histological analyses of the LGNs revealed pronounced shrinkage in cells receiving input from the deprived eye (Wiesel and Hubel 1963a). The second paper in this series (Hubel and Wiesel 1963b) asked whether the distinct cortical response types of the adult cat were present in young kittens deprived of visual experience. In the cortex of visual pattern naive animals (postnatal day P8 before eye opening and P16 with translucent occluders), active cells were hard to find. Stimuli that produced robust responses in adults were either nonexistent or sluggish and fatigued rapidly. Nevertheless, the responding cells were selective for edges, bars, and slits with distinct, though not precisely tuned orientation preferences. Normal binocularity also seemed to have developed in these kittens. Comparison with a P20 normal kitten and a P19 kitten with only one eye covered with a translucent occluder revealed more spontaneous activity, brisker responses to patterned visual stimuli, and stricter orientation preferences than seen in the younger animals. Indeed, these were the responses expected if the cortical circuit was preformed and needed pattern vision only to strengthen young connections.
Disruptive effects of prolonged abnormal visual experience
The third paper in the 1963 series (Wiesel and Hubel, 1963b) grew from these initial developmental observations observations. Earlier, Hubel and Wiesel (1962) published a description of binocular interaction in the mature cat visual cortex. In the third 1963 paper, they introduced the now famous monocular deprivation paradigm and explored the behavior and physiology of the visual cortex of kittens with long-term monocular occlusion. Kittens deprived from 1–4 months of age, when forced to use only the deprived eye, behaved as if they were blind. When the same kittens were allowed to use the non-deprived eye, visual behavior was normal. An adult cat studied with the same paradigm showed no effect of the deprivation. Electrophysiological analyses rapidly revealed the cause of the behavioral blindness: of 84 cells recorded in the cortex contralateral to the deprived eye, only 1 neuron responded to that eye. Normally, an eye dominates responses in the contralateral cortex. However, in an additional experiment where cortical neuron recordings were made contralateral to an eye that had only the thin nictitating membrane sewn across it for 5 months, LGN histology showed no pronounced atrophy in the deprived eye LGN layers. Nevertheless, the ocular dominance histogram showed pronounced non-deprived eye domination. The result was the first indication that the primary locus for the LGN neuron atrophy was in the cortex and not the LGN. Guillery (1972) would later eliminate binocular competition from the nondeprived eye in a small region of cortex and experimentally support the hypothesis that the atrophy in the LGN arose from competition between the eye-specific inputs in the cortex.
On the basis of long-term dark rearing from birth in a variety of mammals, a number of early investigators had argued that in order to develop in the first place, visual connections require visual stimulation (Berger 1900; Goodman 1932). The third 1963 Wiesel and Hubel report (Wiesel and Hubel 1963b) broke with this earlier work by showing abnormal experience at an early developmental stage could permanently disrupt initially formed circuits. This idea was new. Although Konrad Lorenz had introduced the concept of developmental critical periods with his studies of imprinting in young precocial birds (Lorenz 1937), experiments in which normal experience was prevented for short periods in early development did not invariably have long-term effects (Marler and Hamilton 1966). Hubel and Wiesel introduced into the electrophysiological literature an extension of an idea articulated first by Hebb in 1949. They imply in their conclusion to this third report that competition between the two eyes may be occurring, i.e., … “if one eye is not stimulated, the fate of its projections in the central visual pathway may partly depend on whether or not the other eye is stimulated”. Three decades later Antonini and Stryker (1993) finally provided single neuron anatomical evidence of this competition by reconstructing individual LGN axon terminals and demonstrating a highly significant difference between sparse deprived eye terminals and robustly arborized terminals from the nondeprived eye.
Competition in cortical plasticity
In the first of a second series of papers in 1965 Wiesel and Hubel (1965a) directly addressed the question of whether a visually deprived eye's response to deprivation is dependent on the activity of the other eye. They answered it in the affirmative. Both eyes of young kittens were sutured shut, and their striate cortices were recorded from 2.5–4.5 months later. In these animals, most active cells could be driven through both eyes, and some with simple and complex receptive field properties were found. However, many cells lacked a defined orientation preference, and roughly one-quarter of the active cells were unresponsive to visual stimulation. There were no obvious differences between left and right eye layers in the LGN, but comparisons with normal animals revealed some cell size shrinkage in all layers. Furthermore, kittens with this long-term binocular deprivation proved to be behaviorally blind, similar to the dark-reared animals of earlier investigators and to those kittens forced to use a deprived eye after shorter monocular occlusion. Kittens with long-term monocular deprivation were similarly blind in the closed eye, but, in contrast to the binocularly deprived animals, they showed a consistent shift toward dominance by the non-deprived eye, many cells lacked orientation sensitivity, and, as in the binocularly deprived animals, many cells were unresponsive.
The second paper of this second series asked if the activity arriving at the cortex through the two eyes had to be coincident in order to produce cells equally driven by the two eyes. Once again Hubel and Wiesel (1965) answered the question in the affirmative. Kittens raised with surgically induced divergent strabismus for 3 months or 1 year, though behaviorally competent and showing robust normal field properties through either eye, had exceptionally few cells driven by both eyes, and an electrode passing through cortex encountered large areas of cells responding only to inputs from one or the other eye. Similar studies were conducted in animals with alternating monocular occlusion, and the effects on ocular dominance in these animals were more extreme than in the kittens with squint: 91% of the all cortical neurons encountered were driven only by one eye and, as in the squint kittens, there was a pronounced spatial segregation of cells driven by each eye. In fact, this paper gives one of the first indications of ocular dominance columns. Prompted by the eye-dependent segregation seen in the squint and alternating monocular deprivation animals, Hubel and Wiesel went back to examine data in a previous publication focused on orientation columns in normal adult cats (Hubel and Wiesel 1963a). They tabulated the first few units encountered in each penetration to show surface maps of where each eye dominated the cellular responses. Thus this 1965 paper presents the first tangential image of cortical ocular dominance.
In the final publication of the 1965 series, Wiesel and Hubel (1965b) tackled a question that has recently, with the advent of congenital cataract removal and lens replacement in older patients, become highly relevant: namely, the extent of recovery from early visual deprivation (Fine et al. 2003; Ostrovsky et al. 2006). With long-term monocular deprivation, despite subsequent periods where the non-deprived was closed forcing use of the deprived eye, the animals showed little behavioral recovery of vision through the deprived eye, and the cortical neuron response properties remained abnormal. Somewhat different results were found when kittens binocularly deprived for 3 months were kept for over a year with only one eye open. Compared with kittens with only 3 months of binocular deprivation, these cats had many fewer unresponsive cells but also fewer normal cells, and almost twice as many abnormal cells were driven predominantly from the reopened eye. Significantly, most of these cells were abnormal because they lacked orientation sensitivity through the open eye. This result is interesting, in light of recent work, because it suggests that orientation selectivity is significantly more labile to early deprivation than is ocular dominance. Though it may be coincidence, this discordance between ocular dominance and orientation sensitivity is reminiscent of studies indicating that orientation tuning in mammalian visual cortex is sensitive to disruption of NMDA receptor function whereas ocular dominance is not (Fagiolini et al. 2003; Ramoa et al. 2001).
Wiesel and Hubel (1965b) close their discussion of this last paper by listing two possibilities to explain their results: “Either connections once lost are incapable of properly re-establishing themselves, or the failure of recovery may simply be a matter of age”, reflecting something “that happens between the third month and the first year”. In 1965, distinguishing between these two possibilities was impossible, but cell and molecular biology when combined with the appropriate whole animal paradigms promise to reveal what a “matter of age” involves: specifically, which signaling pathways can, and which cannot, change upon vision later in life after early deprivation. Visual developmental plasticity has come a long way since 1963, and most progress has been built upon the experimental procedures first laid out in these six seminal papers. However, we now understand that spontaneous activity in the retina and many brain areas contributes significantly to the refinement of neural circuits frequently long before stimuli from the outside world impact upon the developing brain. We also understand that the critical period for ocular dominance plasticity can be very short, that it depends on inhibitory development, and that there is a latent plasticity even in the mature brain. The latter makes efforts to understand exactly what factors limit this plasticity as the brain grows older ever more important to pursue. The hope is that our new era of genomes, proteomics, and genetic engineering will ultimately return the full visual world to those suffering from complete or partial blindness as a result of disrupted vision early in their life.
- Copyright © 2008 by the American Physiological Society