We tested the concept that lesions of primary visual cortical areas 17 and 18 sustained on the day of birth induce a redistribution of cerebral operations underlying the ability to disengage visual attention and then redirect it to a new location. In cats, these operations are normally highly localizable to posterior middle suprasylvian (pMS) cortex. Three stimulation paradigms were used: (i) movement of a high contrast visual stimulus into the visual field; (ii) illumination of a static light-emitting diode (LED) stimulus; and (iii) a control static auditory stimulus. To test for the redistribution of critical neural operations, cryoloops were implanted bilaterally in the pMS sulcus and in contact with ventral posterior suprasylvian (vPS) cortex. Separate and combined deactivations of pMS and vPS cortices in cats with early lesions of primary visual cortex showed that full, unilateral deactivation of pMS cortex only partially impaired the ability to detect and orient to stimuli moved into the contracooled hemifield. Much more complete impairment required the additional deactivation of ipsilateral vPS cortex. Bilateral pMS deactivation alone, or in combination with bilateral vPS deactivation, largely reversed the unilateral contracooled neglect. For the orienting to static, illuminated LED stimuli, unilateral deactivation of pMS cortex was sufficient to fully impair orienting to stimuli presented in the contracooled hemifield. Bilateral pMS deactivation induced an almost complete visual-field-wide neglect of stimuli. On its own, unilateral deactivation of vPS cortex was without effect on either task, although bilateral vPS deactivations introduced inconsistencies into the performance. Termination of cooling reversed all deficits. Finally, neither the initial lesion of areas 17 and 18 nor cooling of either the MS or vPS cortex alone, or in combination, interfered with orienting to sound stimuli. Overall, our results provide evidence that at least one highly localizable visual function of normal cerebral cortex is remapped across the cortical surface following the early lesion of primary visual cortical areas 17 and 18. Moreover, the redistribution has spread the essential neural operations from the visual parietal cortex to a normally functionally distinct type of cortex in the visual temporal system.
It is now accepted that the young human brain is highly plastic and that lesions interfere with innate development of architecture, patterns of connections and mapped functions, and the lesions trigger modifications in structure, rewirings and representations. In some instances, these modifications are adaptive and result in sparing of functions that would otherwise be lost, or severely handicapped, by equivalent lesions incurred later in life (Milner, 1974; Rudel et al., 1974; Woods, 1981; Ogden, 1989; Stiles and Nass, 1991; Bates, 1999; van den Hout et al., 2000). However, when sparing occurs it may be at a ‘price’ because the spared functions may ‘crowd out’ other functions from their normal regions (Milner, 1974; Teuber, 1975).
Analogous studies on animals foster similar conclusions (Schneider, 1979; Goldman-Rakic et al., 1983; Kolb and Whishaw, 1996; Payne et al., 1996b) and amplify upon the initial examples demonstrating the general plasticity of brain connections (Rakic, 1977; Wiesel, 1982). Moreover, the degree of functional sparing varies systematically with the age at which the cortical damage was incurred, and they show that the sparing is accompanied by anatomically demonstrable changes in brain wiring (Goldman-Rakic et al., 1983; Cornwell et al., 1989; Shupert et al., 1993; Kolb, 1995; Payne et al., 1996b, 2000; Payne and Lomber, 2001). Such studies frequently prompt the question: ‘Does discrete damage of the cerebrum early in life induce a redistribution of spared functions and neuronal operations across the remaining cerebral regions?’ We attempt to answer this question by using temporary cooling deactivation of circumscribed cerebral regions and behavioral analyses of the visually guided behavior of cats that sustained lesions of primary visual cortex shortly after birth; a time when the visual system has substantial capacity to modify its program for developing connections (Payne and Cornwell, 1994; Payne et al., 1996b), the outcome of which can be linked to neuronal compensations (Spear, 1995).
Visual cortex lesions incurred shortly after birth spare the capacity to learn certain visual behavioral tasks involving complex patterns (Cornwell et al., 1989), and spare a number of largely reflexive tasks such as redirection of attention and orienting from a fixation point to a new target (Shupert et al., 1993). Sparing of the latter behavior has an obligatory requirement on the suprasylvian belt of cortex that includes the cortex lining the middle suprasylvian (MS) sulcus and forming the posterior suprasylvian (PS) gyrus (Fig. 1A) (Shupert et al., 1993). These are the same regions that are known, or suspected, to receive substantially expanded projections from other visual structures (Payne et al., 1996b). The involvement of MS and ventral (v) PS cortices is buttressed by electrophysiological studies that show that both regions contain neurons that frequently exhibit well defined receptive field properties and close-to-normal neural activity levels (Doty, 1961; Cornwell et al., 1978; Tumosa et al., 1989; Guido et al., 1990, 1992). However, even with the evidence pointing to posterior (p) MS and vPS cortices as major cerebral players in the sparing of visual functions, no behavioral studies have been carried out to specifically test the actual contributions the two regions each make to spared perceptual and cognitive processing.
Even though we have a reasonable comprehension of the contributions pMS and vPS cortices make to visually guided behavior in intact cats (Lomber et al., 1994, 1996a, b; Lomber and Payne, 1996, 2000a, 2001), we may not accurately anticipate the contributions each region makes to visually guided behavior following early lesions of primary visual cortex. This uncertainty arises because the numerous pathway expansions suggest that functions normally localized to one region may become dispersed following the early lesion. For example, two major efferent systems that may contribute to the redistribution of signals are the expanded projection from vPS cortex to pMS cortex (MacNeil et al., 1996) and the expanded projection from MS cortex that reaches vPS cortex via the superior colliculus and the medial division of the lateral posterior nucleus of the thalamus (Raczkowski and Rosenquist, 1983; Abramson and Chalupa, 1988; Lomber et al., 1995). The presence of these relays suggests that the processing of signals in MS cortex is altered in significant ways by the signals arriving from vPS cortex and vice versa. Moreover, it suggests that one or other region may contribute in novel ways to visually guided behavior. This possibility is the crux of the question posed at the outset, and it is the possibility we test in the current study with two visual detection and orienting tasks, based on moving and static stimuli, and one auditory detection and orienting task. We use the latter as a valuable control task, and performance is anticipated to be unimpaired by experimental manipulations on the visual system. In the current work we also employed the practical and highly economical cooling method to reversibly deactivate and assess the contributions the MS and vPS cortices make to brain function (Goldman and Alexander, 1977; Lomber et al., 1999). This method does not have the attendant drawbacks of the lesion method for assessing brain functions. These limitations have been enumerated elsewhere (Lomber, 1999; Lomber et al., 1999).
Materials and Methods
Subjects and Overview of Experimental Plan
Three domestic female cats and one castrated male cat were used. They were all born in the Laboratory Animal Science Center at the Boston University Medical Center. For all surgical procedures antiseptic precautions were used throughout, and all procedures were approved by the Animal Care and Use Committee of the Boston University Medical Center and carried out in accordance with the policies outlined in the National Institutes of Health Guide for the Care and Use of Laboratory Animals (No. 86-23). Primary visual cortical areas 17 and 18 were removed bilaterally on the day of birth from all four cats (Fig. 1B) (Lomber et al., 1993, 1995), and 6 months later, when the cats had reached adulthood, all four started training on three behavioral tasks requiring detection and orienting to a visual or sound stimulus (Lomber and Payne, 2001). After extensive training on these and other tasks, bilateral pairs of cooling loops were surgically placed (Fig. 1B) within the posterior portion of the middle suprasylvian sulcus (pMS sulcal cortex) and in contact with the ventral portion of the posterior suprasylvian gyrus (vPS cortex) in all four cats using established procedures (Lomber et al., 1996b). The cats were then tested on three detection-and-orienting tasks with and without the cooling loops operational. At the conclusion of all testing, 2-deoxyglucose (2DG) was administered while two, asymmetrically positioned, cooling loops were cooled to experimental temperatures. Following tissue fixation and processing, and preparation of radiograms, regions of low 2DG uptake were identified, the extent of deactivated cortex measured, and the position and extent of the area 17/18 ablation verified.
Behavior Apparatus, Training and Testing
Two arenas were employed for testing the cats' abilities to detect and orient to visual and auditory cues.
Standard Arena (Moving Visual Stimulus)
The standard apparatus is designed to test the cat's ability to detect and orient to a visual stimulus suddenly moved into the visual field through an aperture located on or below the horizontal meridian (Lomber et al., 1996b). The apparatus is a white semicircular platform with a diameter of 88 cm and a perimeter wall 30 cm high. At the base of the wall and around the perimeter are thirteen 3 × 4 cm openings. The openings are separated by distances that subtend an angle of 15° relative to the center of the arena. Openings are positioned from left 90° to right 90°. Background illumination was provided by standard room illumination and was at photopic levels.
Visual and Auditory Arena (Static Visual and Auditory Stimuli)
The overall dimensions of the visual and auditory apparatus are similar to the standard perimetry arena. The significant differences are that the perimeter wall is 45 cm high and has positioned on it, at a height of 25 cm, 13 light emitting diode (LED)–microspeaker complexes that are separated by distances that subtend an angle of 15° relative to the center of the arena. Twelve peripheral LEDs, model T-1, and one central LED, model T-1¾, are illuminated using a 2 V DC current source. Background lighting was reduced and set at low photopic levels. The peripheral speakers emitted broadband white noise 2 dB above background (~52dB) and the central speaker emitted noise at 5 dB above background. The speakers and LEDs were connected to a switchboard located behind the arena. Both visual and auditory stimuli were maintained for the duration of each trial. Below each LED–speaker complex is a 3.5 cm diameter opening for the delivery of food rewards. This arena has many features in common with the arena described elsewhere (Stein et al., 1989).
Behavioral training began in adulthood and was in all instances binocular. Standard procedures were adopted (Lomber and Payne, 2001) and included a predetermined, balanced, yet non-systematic, sequence for stimulus presentation. Cats progressed from the moving-visual, onto static-visual and finally onto auditory training only after asymptotic performance was reached on the prior task. Blank trials were incorporated to guard against and measure frequency of ‘scanning’ of the visual and auditory fields.
All testing was performed under binocular viewing conditions, and a testing block consisted of 28 trials (two presentations at each of the 12 peripheral locations and four presentations at the central position). Testing sessions were typically carried out once per day, with 10–15 blocks of trials being collected in each session. All six combinations of loci deactivations were carried out but not all in the same session. They included: (i) pMS unilaterally, (ii) pMS bilaterally, (iii) vPS unilaterally, (iv) vPS bilaterally, (v) pMS and vPS unilaterally, and (vi) pMS and vPS bilaterally.
Cooling loops were chilled to 3°C, a temperature known to deactivate the full thickness of cortex adjacent to the cooling loop (Payne and Lomber, 1999; Lomber and Payne, 2000a, b). For each locus, or pair of loci, being tested a five-step paradigm was followed: (i) baseline data was collected with all sites active. (ii) Testing began while either the left pMS or left vPS alone, or left pMS and vPS regions together, were cooled and deactivated. (iii) Cooling of the homotopic site(s) in the right hemisphere was/were then added to permit the effects of bilateral deactivation to be examined. (iv) The cooling of the left site(s) was terminated and cortex released from cooling while the site(s) in the right hemisphere remained deactivated. (v) Baseline levels were reestablished after release of the right site(s) from cooling. The procedure was repeated at least six times (three times in the order given above above and three times in reverse order). Procedures were repeated at regular intervals over several months to test the stability of the cooling impact. Neither attenuation nor exaggeration of cooling-induced deficits suggest that either cooling-induced neural compensations or cooling damage was detected. In total, >24 000 testing trials were carried out.
After a 9–10 month period of testing, each cat was acclimated over a 2–3 day period to a cat-restraining bag. After acclimation, two cooling loops (one pMS and one vPS in opposite hemispheres) were chilled to the experimental temperature of 3°C and the cat was injected i.v with four doses of 2-deoxy-d-[U-14C]glucose (25 μCi/kg) at 5 min intervals as described previously (Payne and Lomber, 1999). Following a final 15 min uptake period, 2000 units of Heparin®, 2 ml of 1% sodium nitrite and 35 mg/kg of sodium pentobarbital were injected i.v. in that order. Once deep anesthesia had been reached, the brain was fixed by arterial perfusion with a phosphate buffered paraformaldehyde solution containing 4% sucrose. This method of euthanasia by exsanguination is consistent with the recommendations of the American Veterinary Medical Association Panel on Euthanasia (Andrews et al., 1993) and for the preparation of scientifically useful tissue. The fixed brain was removed, photographed, blocked, coated in egg albumin and frozen in methylbutane at –35°C. Thirty minutes later, the brain was transferred to a –80°C freezer until it was sectioned.
Tissue Processing and Examination
Histology and Radiography
Twenty-six-micrometer-thick coronal, frozen sections were cut and mounted onto coverslips affixed to cards. Sections were placed against Agfa-Gevaert Structurix X-ray film and stored at –80°C in light-tight cassettes for 40–60 days (Payne and Lomber, 1999), when they were developed with Kodak D-19B developer. Adjacent sections were stained for Nissl substance or cytochrome oxidase to verify location of the cooling loop and absence of surgical or cooling induced damage. The same sections were also used to estimate the extent and location of the primary visual cortex lesion.
Extent of Cooling Deactivation
Cooling-induced decreases in 2DG uptake are obvious (Payne et al., 1996c) and only require sophisticated imaging equipment to assay the gradients on the fringes of the deactivation. For these purposes we used an AIS™ microcomputer-based image analyzer (Imaging Research, Inc., St Catherines, Ontario, Canada), in conjunction with 14C standards (Amersham Corp., Arlington Heights, IL) and calibration curves (Gonzalez-Lima, 1992).
The size and location of the visual cortex ablation was assessed directly and by using standard procedures to examine cerebral cortex and LGN in the coronal sections stained for Nissl substance and cytochrome oxidase reaction product (Payne and Lomber, 1996).
For each animal, initial within subject t-tests were carried out between the pre- and post-cool active conditions to confirm stationarity of behavior across the group of test blocks. Individual means during the active, deactivated and ‘reactivated' conditions were calculated for each testing session. For all the histograms presented, group means and standard errors are based upon individual means from equal numbers of testing sessions for each cat. Data for the 0° position in the static and auditory arena were divided equally between the two hemifields because in this instance the cynosure was also the stimulus. This step was unnecessary for the moving stimulus because the stimulus was introduced into the arena adjacent to the cynosure. The mean performance while the cortex was either warm or cold was compared for all subjects examined on a given task using within-subject t-tests. Within the Results section, unless otherwise stated, if a difference is described as being significant or reliable, it has a P ≤ 0.01.
Evaluation of Cortical Ablations and Thalamic Degeneration
Gross examination of the fixed brain showed that nearly all of the marginal and posterolateral gyri, as well as the medial aspect of the junction of the middle–posterior suprasylvian gyri, were removed from every hemisphere. Figure 2 shows reconstructions of the cortical ablations plotted onto standardized drawings of the dorsal view of the cerebral cortex. Microscopical analyses also revealed minor, variable degrees of tissue sparing on the upper bank of the splenial sulcus and caudally on the posterolateral gyrus. These regions can be readily identified on the coronal sections taken through regions containing the ablation (Fig. 2). The extent of the ablation of areas 17 and 18 was verified by examination of the dorsal lateral geniculate nucleus (dLGN). Severe retrograde degeneration was evident throughout layers A and A1 of the dLGN, although a small number of neurons survived in these layers, as is typical following early lesions of primary visual cortical areas 17 and 18 (Doty, 1961; Cornwell et al., 1978; Spear et al., 1980; Lomber et al., 1995; Payne and Lomber, 1998). In addition, there was minor to moderate degeneration evident in the medial interlaminar nucleus and the C-laminae. Degeneration was not detected in either the pulvinar or lateral posterior nuclei. These patterns of degeneration are consistent with both our earlier descriptions (Lomber et al., 1993, 1995) and the known pattern of anatomical connections between dLGN and visual cortex at the time the lesion was made (Cornwell et al., 1984; Payne et al., 1988). Overall, between 93 and 97% of combined areas 17 and 18 were removed from each hemisphere. In addition, there was moderate to major inclusion of area 19 and inclusion of portions of the splenial visual area (Kalia and Whitteridge, 1973) in the ablation. The inclusion of these regions was a natural product of our attempts to maximize the ablation of areas 17 and 18.
Structure of Cortex Beneath the Cryoloops and Extent of Cooling Deactivation
Examination of Nissl and myelin tissue revealed that neither the surgical procedures nor the chronic presence of the cooling loops and the repeated deactivations had any impact on the structural integrity of the MS and vPS cortices adjacent to the cryoloops. This conclusion was extended to long-term functional integrity in both the MS and vPS regions because cytochrome oxidase reaction product was of a high or medium level, respectively, as is typical of brains with early lesions of primary visual cortex (Long et al., 1996). Concordant with this view, the 2DG radiograms of cortex adjacent to the non-operational MS and vPS cryoloops during 2DG administration exhibited rich gray densities indicative of normal 2DG, uptake and normal brain function at the time of the 2DG administration (Fig. 3, left). They contrasted with the very pale regions indicative of greatly reduced 2DG uptake, and absence of neural activity in the MS and vPS cortices adjacent to the two cryoloops that were operational at the time the 2DG was administered (Fig. 3, right).
In the pMS sulcal region of the cortex, low 2DG uptake characterized both the medial and lateral banks of the pMS sulcus, and was slightly asymmetric in favor of the medial direction (Fig. 3, upper right). This asymmetry reflects the overall direction of blood flow from lateral to medial in the region, which contains the lateral spread of cooling but exports the cooling at short distance in the medial direction. The deactivated region extended on both banks from approximately coronal level P3 to A8 and includes the regions labeled by Palmer et al. (Palmer et al., 1978) as PMLS and PLLS, dorsal DLS, posterior AMLS and posterior ALLS, and labeled areas LS (Sherk, 1986; Grant and Shipp 1991) and PEV (Grant and Shipp, 1991) in the intact brain.
In the vPS region, low 2DG uptake characterized all of the vPS cortex (Fig. 3 lower right) and much of the hippocampal fusiform gyrus, as intended. This combined region corresponds to virtually all of area 20 on the lateral and tentorial surfaces (Updyke, 1986) and the ventral-most portion of area 21 (Tusa and Palmer, 1980). There was little or no spread of cooling to area PS of Updyke (Updyke, 1986).
Visual Detection and Orienting Task: Moving Stimulus
After more than 2000 training trials, cats with ablations of areas 17 and 18 incurred on the day of birth are extremely proficient at detecting and orienting towards a high contrast stimulus moved into the perimetry arena through any of the arena-floor openings [Fig. 4A(i)]. This level of performance approaches that of intact cats in other studies (Shupert et al., 1993; Lomber and Payne, 1996), and is far superior to the 30–60% level of proficiency exhibited by cats that incurred equivalent, or smaller, lesions in adulthood (Shupert et al., 1993; Payne and Lomber, 2001).
pMS Sulcal Cortex Deactivation
Deactivation of either left or right pMS sulcal cortex [Fig. 5A(ii,iv)] resulted in a mean decrease of 34% in orienting proficiency to stimuli moved into the contracooled hemifield (opposite side to the deactivation). The impairment was spread across the hemifield with more peripheral positions exhibiting a greater fragility [Fig. 4B(i)]. Interleaved presentation of stimuli in the ipsicooled hemifield (same side as the deactivation) revealed virtually no impact of the cooling on orienting performance [Fig. 5A(ii,iv)]. As we have learned to expect from the intact brain, additional deactivation of the homotopic pMS site in the opposite hemisphere restored visual orienting into the previously partially neglected hemifield [Fig. 5A(iii)] for all positions examined [Fig. 4B(ii)]. Subsequent release from all cooling resulted in levels of orienting proficiency that matched precool levels [Figs 4A(ii) and 5A(v)].
vPS Cortex Deactivation
Unilateral deactivation of vPS cortex did not modify the cat's ability to detect and orient towards stimuli moved into either the contracooled or ipsicooled hemifields [Figs 4C(i) and 5B(ii,iv)]. However, bilateral deactivation of vPS cortex cooling significantly reduced orienting throughout the visual field from 94 ± 2 to 77 ± 2%, and this reduction was manifest with only a slightly greater decrease for peripheral than for more central positions in the visual field [Fig. 4C(ii)]. Following termination of all cooling, orienting performance returned to precool levels [Figs 4A(ii) and 5B(v)].
Moving Visual Stimulus: Deactivation of both pMS and vPS cortices
The partial impairments induced by unilateral pMS and bilateral vPS deactivations suggested to us that both regions in cats with lesions of areas 17 and 18 sustained on postnatal day 1 (day of birth; P1) contribute in important ways, but under different circumstances, to proficient detection and orienting toward visual stimuli. Based on these observations, we tested for, and showed, the additive impact of combined unilateral cooling of pMS and vPS cortices [Fig. 6A(iii)] on orienting performance, which was reduced to 18.9 ± 4.3% and approached a full neglect of that hemifield. Interleaved trials showed that orienting into the ipsicooled hemifield was unimpaired.
The low level of performance induced by the combined pMS and vPS deactivations was reversed either rapidly by terminating cooling [Fig. 6A(iv)] or gradually by additional cooling of first the contralateral pMS cryoloop [Fig. 6B(ii)], and then more completely with the additional cooling of the contralateral vPS cryoloop [Fig. 6B(iii)]. However, the gradual reversal was incomplete and only reached 82%, because there was a maintained minor impairment in orienting to stimuli presented in the right visual field which was matched by an additional novel impairment in the previously fully functional left, hemifield ipsilateral to the initial cooling. Finally, termination of cooling on one side (pMS and vPS) resulted in reinstatement of a substantial neglect of the hemifield contralateral to the cooling [Fig. 6B(iv)], and termination of all cooling resulted in a return of highly proficient orienting (>94%).
Figure 7 shows the impact of these sequential coolings relative to visual field position. Unilateral cooling of pMS cortex reduced orienting proficiency across all parts of the contracooled hemifield with the peripheral-to-center asymmetry of impact noted earlier (Fig. 7B). The addition of cooling of vPS cortex on the same side substantially diminished the poor to modest orienting performance further, and modest orienting performance was only maintained for the contralateral juxta-central position. The further addition of cooling of pMS cortex in the opposite hemisphere resulted in improved, but less than perfect, orienting proficiency throughout the previously neglected hemifield (Fig. 7D). Auxiliary cooling of the contralateral vPS cortex improved orienting proficiency further (Fig. 7E), and it reached close to perfect levels for right juxta-central, 15° and 30° positions, but partial impairments remained at more peripheral positions. Concurrent with this improvement, a deficit in orienting performance was detected in the periphery of the initially robust left hemifield ipsilateral to the initial unilateral right coolings. Finally, cessation of all cooling completely reversed the impairments (not shown).
Visual Detection and Orienting Task: Static Stimulus
Overall, cats with lesions of areas 17 and 18 sustained on P1 orient less proficiently to illuminated, high contrast LEDs in a darkened arena [65 ± 4%; Fig. 8A(i)] than to a moving stimulus in a bright arena [94 ± 3%; Fig. 4A(i)], and comparison of radial diagrams reveals that the weakness in orienting is towards LEDs illuminated in the peripheral one-third of the field from 60° outwards [Fig. 9A(i)], orienting to more central positions being close-to-perfect or perfect [Fig. 9A(i)]. Since performance is inferior to that of highly trained intact cats, and not markedly better than that exhibited by adult cats with equivalent lesions of primary visual cortex (Payne et al., 2000; Payne and Lomber, 2001), performance represents residual visual capabilities rather than spared vision. Visual capacities after lesions of primary visual cortex can be divided into three main categories: (i) residual vision describes the visual capacities that remain after lesions; (ii) recovered vision describes the visual capacities that remain after lesions if these capacities emerged from, and are superior to, to residual vision; recovery is rarely complete; and (iii) spared vision describes the visual capacities that are present after lesions incurred in the earlier part of life before faculties have fully matured; spared vision is always greater than both residual and recovered vision.
pMS Sulcal Cortex Deactivation
Unilateral deactivation of pMS sulcal cortex induced a profound neglect of the LEDs illuminated in the contracooled hemifield that was initially complete and more profound [10.1 ± 5.7%; Fig. 8A(ii,iv)] than the neglect of the moving stimulus [59 ± 4%; Fig. 4A(ii,iv)]. In contradistinction, no negative impact of the cooling on orienting responses into the ipsicooled hemifield was identified [Fig. 8A(ii)], and the radial plot was largely indistinguishable from the data collected during the precool trials [Fig. 9B(i)], although there was a hint of a slight improvement in orienting into the ipsicooled hemifield [compare Fig. 8A(ii,iv) with Fig. 8A(i)].
In contrast to bilateral deactivation of pMS cortex and the test with the moving stimulus, which resulted in unimpaired orienting, the same bilateral deactivation of pMS cortices induced a substantial neglect of the illuminated LEDs in both hemifields [Fig. 8A(iii)]. The radial plots reveal that the limited orienting capability that remained was only for central-most positions [Fig. 9B(ii)]. Terminations of cooling reversed the cooling-induced deficits and performance duplicated precool levels.
vPS Cortex Deactivation
Unilateral deactivation of vPS cortex did not impair orienting to an illuminated LED, performance was indistinguishable for targets illuminated in the ipsicooled and contracooled hemifields [Fig. 8B(ii,iv)], and the radial diagrams are indistinguishable from those representing data collected prior to cooling [compare Fig. 9C(i) and Fig. 9A(i)]. In contrast, bilateral deactivation of vPS cortex impaired orienting to LEDs illuminated in both hemifields [Fig. 8B(iii)], and residual orienting capability was focused solely on the cynosure and flanking regions [Fig. 9C(ii)]. In all instances, orienting performance following cessation of cooling reproduced the performance measured prior to cooling [Figs 8B(v) and 9A(ii)].
Auditory Detection and Orienting Task
Overall, cats with ablations of primary visual cortical areas 17 and 18 sustained on the day of birth are moderately proficient (65–75% correct) at orienting to a stationary, broadband white-noise auditory stimulus [Fig. 10A(i)], and radial diagrams reveal differential performance levels for central versus more peripheral positions. Orienting to the auditory stimulus is perfect, or nearly so, for positions extending out to 45° left and right, with poorer or no orienting performance to more peripheral locations (Fig. 11A). The characteristics of this performance are comparable to those exhibited by intact cats and cats that incurred equivalent lesions of primary visual cortex sustained in adulthood (Payne et al., 2000; Payne and Lomber, 2001).
The upper cells in Table 1 summarize the known capabilities and performance of intact cats and cats with lesions of areas 17 and 18 incurred in adulthood or on P1. The superior performance of the P1-lesioned compared with the adult-lesioned group provides the evidence for sparing of the orienting task that employs a moving visual stimulus. The lower cells summarize the results of the current study on P1-lesioned cats during cooling deactivation of the MS and vPS cortices either alone or in combination. They show that full, unilateral deactivation of pMS cortex only partially impaired the almost fully spared ability to detect and orient to stimuli moved into the contracooled hemifield. Much more complete impairment required the additional deactivation of the ipsilateral vPS cortex. Bilateral pMS deactivation alone or in combination with bilateral vPS deactivation largely reversed the unilateral contracooled neglect. For the orienting to static illuminated LED stimuli, unilateral deactivation of pMS cortex was sufficient to fully impair orienting to stimuli presented in the contracooled hemifield. Bilateral pMS deactivation induced an almost complete visual-field-wide neglect of stimuli. For the already degraded orienting to static LED stimuli illuminated in the visual field, unilateral deactivation of pMS cortex was sufficient to fully impair orienting to stimuli presented in the contracooled hemifield. On its own, unilateral deactivation of vPS cortex was without effect on either task, although bilateral vPS deactivations introduce inconsistencies into the performance. Termination of cooling reversed the deficits. Finally, cooling of neither the MS nor vPS cortex alone, or in combination, interfered with orienting to sound stimuli.
It has been established in a comprehensive series of studies that removal of primary visual cortical areas 17 and 18 from kittens on the day of birth, or shortly thereafter, induces a reproducible multi-component rewiring of all or most of the remaining visual system (Payne and Cornwell, 1994; Payne et al., 1996b). The rewirings include systematic degenerations of neurons normally richly connected with the ablated cortex, and systematic expansions of pathways that bypass both the removed cortex and distant regions exhibiting degeneration. The degenerations and expansions include all major divisions of the visual system, from the retina, through the thalamus, regions of extrastriate cortex and superior colliculus, and they are accompanied by known neural compensations in the LGN, middle suprasylvian cortex and superior colliculus (Spear et al., 1980; Tumosa et al., 1989; Guido et al., 1990; Mendola and Payne, 1993). It is believed that both the anatomical adjustments and the neural compensations contribute significantly to the sparing of learned and reflexive visually guided behaviors linked to suprasylvian cortex and normally associated with visual streams reaching temporal and parietal cortices, respectively. However, no tests of this proposition have been carried out until now.
One of the most robust pathway expansions is from the retina via the surviving neurons in the LGN to the middle suprasylvian cortex in the visual parietal region (Tong et al., 1984; Kalil et al., 1991; Lomber et al., 1993, 1995; Payne and Lomber, 1998), where some of the most robust neural compensations have been identified, including sparing of at least some aspects of motion processing amongst other properties (Spear et al., 1980; Guido et al., 1990). The identified compensations also likely rely heavily upon signals transmitted from the retina through the midbrain to the medial part of the lateral posterior complex, which is induced by primary visual cortex to double its already massive projection to the MS cortex, and upon expanded projections from vPS cortex (Lomber et al., 1995; MacNeil et al., 1996). However, the physiological studies only provide corollary data, and do not test directly for the contributions specific structures make to the sparing and compensations. Even so, in accord with the physiological data, cats with early lesions of primary visual cortex exhibit sparing of visually guided orienting to stimuli moved into an arena but not to illumination of LEDs, although substantial ability to orient to the latter stimuli remains following the lesion (Shupert et al., 1993; Payne et al., 2000; Payne and Lomber, 2001).
From studies using cooling deactivation to probe the functions of the cerebral cortex in normal cats, these same two visually guided behaviors are critically dependent upon pMS cortex, although the superior colliculus is also a critically dependent structure (Lomber and Payne, 1996, 2001; Lomber et al., 1996b; Payne et al., 1996a). Even so, the dependence on pMS cortex is task dependent because unilateral, but not bilateral, deactivations of it impair orienting to a stimulus moved into the contra-cooled hemifield, whereas for the illumination of the static LEDs orienting into the contracooled hemifield is blocked by either unilateral or bilateral deactivation (Lomber and Payne, 2001). In terms of the deficits induced by unilateral and bilateral deactivations on the moved and static versions of the detection and orienting tasks, the dependencies established for the intact cat are also represented in cats with early lesions of primary visual cortical areas, but only partially so because there are some notable differences. These differences are summarized and can be compared in Table 1 (lower cells).
Both pMS and vPS cortices of cats that incurred lesions of areas 17 and 18 on the day of birth contribute importantly to the sparing of detection and orienting performance toward moved targets. However, the respective contributions vary in magnitude and become apparent under different experimental circumstances. For example, even though we are confident from our 2DG measures that the full extent of pMS cortex was deactivated by the cooling, we could only induce dependable (meaning present in all four cats) inconsistencies in orienting performance to the moved stimulus when the deactivation was unilateral. These inconsistencies are in marked contrast to the profound, virtually complete, neglect induced by equivalent cooling deactivation of the intact cat. We thus have a result that demonstrates a muted contribution of pMS cortex to spared orienting, and one that argues forcefully for contributions to spared orienting performance by regions that were not deactivated. At least one of these regions is vPS cortex, because addition of its deactivation to an already existing unilateral pMS deactivation exaggerates the impairment in performance and induces a neglect that, in its profundity, approaches the neglect identified during unilateral cooling of pMS cortex alone in intact cats. However, unilateral deactivation of vPS cortex alone has no impact. Thus, there is a lessened role for pMS cortex in visual orienting and a heightened involvement of vPS cortex. This conclusion promotes the view that the circuitry underlying the sparing of visually guided detection and orienting behavior in response to moving stimuli has been modified by the early lesion of primary visual cortical areas 17 and 18, and it has become redistributed across the cortical surface from its normally highly localizable site on the banks of the pMS sulcus to include vPS cortex. This view is buttressed by the results of bilateral cooling deactivation of vPS cortices that induce global inconsistencies across the visual field in orienting performance, but no profound neglect. It is important to state that we are confident that the deficit induced by vPS deactivations is one of a failure to disengage and redirect attention and not one of blindness or some other global visual impairment, because as yet unpublished companion studies of ours show that the same cats are able to reliably make certain types of basic visual form discriminations when vPS cortex is deactivated bilaterally.
At present, we do not know the basis for the broad, modest decrease in orienting behavior during bilateral cooling deactivation of vPS cortex in the absence of any impact during unilateral vPS deactivation. We can only speculate that each vPS cortex contains a representation of all, or almost all, of the visual field and both representations need to be silenced to induce a weakness in orienting. Such, suprahemifield representations have not been identified in visual field maps of the temporal cortex in intact cats (Tusa and Palmer, 1980; Payne, 1993), although they have been identified in monkeys (Gross, 1973a, b).
Even with the rewiring and redistribution of functions, incremental addition of deactivation of first pMS cortex and then vPS cortex in the hemisphere opposite a combined pMS and vPS deactivation first partially, and then completely, restores orienting ability to moving stimuli presented across large regions of the visual field. This result parallels observations made in intact cats when pMS cortex is deactivated first unilaterally and then bilaterally (Lomber and Payne, 1996). However, the restoration in the central two-thirds of the previously neglected hemifield in the early-lesioned cats is at a cost of orienting to the peripheral one-third of the previously responsive field. These results suggest that pMS and vPS cortices are critical to orienting behavior only under certain conditions, and that other structures, such as the superior colliculus (Lomber and Payne, 1996; Lomber et al., 1996b; Payne et al., 1996a), are sufficient to support orienting to stimuli in the central two-thirds, but not the peripheral one-third, of the visual field when the major cortical structures are silenced bilaterally. The basis for the functional restoration in orienting supported by the bilateral deactivations is thought to be intimately related to a restoration in the balance of the competitive mechanisms operating to allow attentional neural machinery to gain control over motor responses (Lomber and Payne, 1996; Lomber et al., 1996b).
Largely concordant results and conclusions were obtained for the residual detection and orienting visual capacities tested by the illumination of LEDs. However, with this stimulus, as in intact cats, bilateral deactivations even of pMS cortices virtually abolished orienting throughout the visual field, and not restoration of orienting ability. These observations suggest that the competitive mechanisms enunciated to account for the differences in the results of unilateral and bilateral cooling when cats were tested with moving stimuli do not apply when stationary LED stimuli are used. These results also argue compellingly that pMS and vPS cortices alone are critically important structures that contribute to orienting to stationary LED stimuli, and no other structures can subsume their roles. Reassuringly, neither the lesion of primary visual cortex nor any of the deactivations in visual parietal and temporal regions had any impact on the detection and orienting of sound stimuli, thus verifying the specificity of the lesions and deactivations for visual functions, and that the deactivations had no direct impact on the motor reporting system.
An important element in comprehending the differences in the conclusions reached for the moving and static stimuli may be that a moving stimulus has a greater potency to activate neurons. This greater potency may be based on the spatial translations and the larger ensembles of neurons that are activated compared with illumination of the static LED stimulus at a single visual coordinate. The greater potency of the moved stimulus may also permit us to detect the visual sparing above residual visual function, whereas neural activity generated by the static stimulus may be insufficient to permit such a separation.
Overall, our results provide a convincing, affirmative answer to the question posed in the introduction by showing that at least one of the highly localizable functions of normal cerebral cortex has been remapped across the cortical surface as an outcome of the early lesion of primary visual cortical areas 17 and 18. Moreover, the redistribution has spread the essential neural operations from visual parietal cortex to a normally functionally distinct type of cortex in the visual temporal system.
Finally, what we have not be able to glean from our data is whether there is a ‘price’ to pay for the redistribution, and that functions normally associated with vPS cortex are displaced, or ‘crowded out’ (Teuber, 1975), as the region acquires or emphasizes functions normally only weakly represented, if at all. We also do not know what functions may have been acquired by pMS cortex through pathway rewirings and the diminution of contributions to detection and orienting functions. Hopefully, these will be topics of future reports.
We thank Christopher Gelston, Helena Madriason and Shana O'Mara for assistance with the training and testing of the animals. We also thank Claus Hilgetag, Jarrett Rushmore and Peter Stiers for their comments on an earlier version of the manuscript. The work was supported by grants from the National Institute for Neurological Disorders and Stroke (#33975) and the National Institute of Mental Health (#44647).
Address correspondence to Stephen G. Lomber, Laboratory for Visual Perception and Cognition, Department of Anatomy and Neurobiology, Boston University School of Medicine, 700 Albany Street, Boston, MA 02118, USA. Email: email@example.com.
|Moving visual||Static visual||Auditory|
|Upper cells: +++++ = highly proficient performance; ++ = modest performance. Lower cells: 0 = no change; ↓ = minor to moderate decrease; ↓↓ = large decrease; – = not tested.|
|P1||pMS — uni||↓||↓↓||0|
|pMS — bi||0||↓↓||0|
|vPS — uni||0||0||0|
|vPS — bi||↓||↓||0|
|MS and vPS — uni||↓↓||–||–|
|MS and vPS — bi||0||–||–|
|Moving visual||Static visual||Auditory|
|Upper cells: +++++ = highly proficient performance; ++ = modest performance. Lower cells: 0 = no change; ↓ = minor to moderate decrease; ↓↓ = large decrease; – = not tested.|
|P1||pMS — uni||↓||↓↓||0|
|pMS — bi||0||↓↓||0|
|vPS — uni||0||0||0|
|vPS — bi||↓||↓||0|
|MS and vPS — uni||↓↓||–||–|
|MS and vPS — bi||0||–||–|