J Physiol Volume 513, Number 2, 507-519, December 1, 1998
The Journal of Physiology (1998), 513.2, pp. 507-519
© Copyright 1998 The Physiological Society
Visual responses of neurones in the second visual area of flying foxes (Pteropus poliocephalus) after lesions of striate cortex
Agnes P. Funk and Marcello G. P. Rosa
Vision, Touch & Hearing Research Centre, Department of Physiology and Pharmacology, The University of Queensland, QLD 4072, Australia
Received 31 March 1998; accepted after revision 18 August 1998.
 |
ABSTRACT |
- The first (V1) and second (V2) cortical visual areas exist in all mammals. However, the functional relationship between these areas varies between species. While in monkeys the responses of V2 cells depend on inputs from V1, in all non-primates studied so far V2 cells largely retain responsiveness to photic stimuli after destruction of V1.
- We studied the visual responsiveness of neurones in V2 of flying foxes after total or partial lesions of the primary visual cortex (V1). The main finding was that visual responses can be evoked in the region of V2 corresponding, in visuotopic co-ordinates, to the lesioned portion of V1 ('lesion projection zone'; LPZ).
- The visuotopic organization of V2 was not altered by V1 lesions.
- The proportion of neurones with strong visual responses was significantly lower within the LPZs (31·5 %) than outside these zones, or in non-lesioned control hemispheres ( > 70 %). LPZ cells showed weak direction and orientation bias, and responded consistently only at low spatial and temporal frequencies.
- The data demonstrate that the functional relationship between V1 and V2 of flying foxes resembles that observed in non-primate mammals. This observation contrasts with the 'primate-like' characteristics of the flying fox visual system reported by previous studies.
|
INTRODUCTION |
The primary visual area (V1) is an homologous area present in all mammalian species studied so far. Cells in V1 receive their main thalamic input from the dorsal lateral geniculate nucleus, (LGNd; Garey et al. 1991), and in each hemisphere, form a precise representation of the entire contralateral visual field (e.g. Rosa et al. 1993). In most, if not all, mammals V1 is bordered by a second visual area (V2), which receives thalamic inputs primarily from the pulvinar-lateral posterior complex, in addition to a variable amount of input from the LGNd (Garey et al. 1991). Like V1, area V2 forms a complete representation of the visual field, with the vertical meridian of the visual field being represented at the border between these areas (e.g. Rosa et al. 1994).
One of the questions raised by this organization is whether V1 and V2 function in parallel, or whether they form a hierarchical series, in which visual responses in one area depend upon inputs from the other. Studies in primates have concluded that destruction or reversible inactivation of V1 renders neurones in V2 unresponsive to visual stimulation (Cowey, 1964; Schiller & Malpeli, 1977; Girard & Bullier, 1989), whereas inactivation of V2 has a comparatively minor effect upon V1 neurones (Sandell & Schiller, 1982). In contrast, experiments in non-primates have revealed that neurones in area V2 remain responsive to photic stimuli after inactivation or destruction of V1 (cat: Dreher & Cottee, 1975; rat: Olavarria & Torrealba, 1978), despite quantitative changes in their response properties (Dreher & Cottee, 1975; Sherk, 1978; Molotchnikoff & Hubert, 1990; Chabli et al. 1998). Thus, the visual pathway of primates may be unique in that V1 forms a critical link between the thalamus and V2. However, more species need to be studied to confirm the validity of this distinction.
In the present study, we explored the responses of V2 neurones after destroying large portions of V1 in grey-headed flying foxes (Pteropus poliocephalus). Flying foxes have been at the centre of a controversy regarding evolutionary relationships since Pettigrew (1986) found that megachiropterans and primates share a specialized pattern of visual field representation in the superior colliculus, as well as several other neuroanatomical characteristics. This led to his hypothesis that flying foxes form a sister group of the Order Primates (Pettigrew et al. 1989; Rosa & Schmid, 1994), with the corollary that chiropterans are diphyletic. We were therefore interested in establishing whether or not lesions of V1 in these animals would suppress the responsiveness of V2 cells, a result that would add to the list of 'primate-like' characteristics of the flying fox visual system.
|
METHODS |
Four adult flying foxes (Pteropus poliocephalus), weighing 670-860 g, were used. In three cases a lesion of V1 was created via subpial aspiration, while one animal served as a non-lesioned control. Electrophysiological recordings were made 11-16 days post-lesion. All experiments were conducted following the ethical guidelines established by the National Health and Medical Research Council of Australia, and were authorized and monitored by the University of Queensland's Animal Experimentation Ethics Committee.
Cortical lesions
The animals were anaesthetized with intramuscular injections of ketamine (50 mg kg-1) combined with xylazine (3 mg kg-1). Additional doses of ketamine (10-20 mg) were used to maintain a surgical level of anaesthesia (evaluated by monitoring the leg withdrawal and corneal reflexes). Injections of dexamethasone (0·4 mg kg-1 I.M.) and atropine (0·15 mg kg-1, I.M.) were also administered. Under sterile conditions, craniotomies 7-10 mm in diameter were made over the right hemisphere, allowing access to V1 and V2. After removal of the dura mater, a suction probe was used to destroy V1, using previously published maps of this area as a guide (Rosa et al. 1993). The left hemisphere was left untouched. The exposed parts of the cortex and brainstem were protected with a sterile soft contact lens, and the piece of skull removed during surgery was cemented back in place with bone wax and dental acrylic. Special care was taken to ensure a smooth acrylic covering, without imperfections or cutting edges that could cause discomfort. Once the acrylic was dry, the muscle and skin were sutured back in anatomical layers. Injections of long-lasting antibiotics (Norocillin, 0·2 ml, I.M.) and saline with glucose (5 ml, subcutaneous, to prevent dehydration during the recovery period) were also administered. The flying foxes were then placed in a dimly lit, warm room, under the close supervision of one of the experimenters. They recovered their normal posture within 5 h of the surgery, and were able to eat and drink shortly thereafter.
Electrophysiological recordings: animal care and preparation
Each of the flying foxes underwent a single recording session. The procedures for anaesthetic induction and medication were the same as those described for the lesion surgery. A tracheotomy was performed, and a tracheal tube (3·0-3·5 mm in diameter) inserted to enable artificial ventilation. The animal was placed on a thermostatically controlled heating pad, and its head was positioned in a stereotaxic frame. The cortex was exposed, and an acrylic well was constructed around the craniotomy, being secured to the skull by orthopaedic screws. A rod attached to an adjustable arm (mounted on the stereotaxic frame) was positioned over the frontal mid-line, and fixed to the acrylic well. This arrangement allowed the head to be supported without the need for stereotaxic bars, and offered an unhindered field of vision. The well was then filled with silicone oil, and a picture of the cortical surface was taken for plotting of electrode penetration sites.
After all surgical procedures were finished, the animal was administered an intravenous infusion of a mixture of pancuronium bromide (0·15 mg kg-1, followed by 0·15 mg kg-1 h-1), sufentanil (10-15 µg kg-1 h-1), ketamine (6 mg kg-1 h-1) and dexamethasone (0·4 mg kg-1 h-1), in a saline-glucose solution. This induced muscular paralysis while maintaining anaesthesia. The animal was artificially ventilated with a gaseous mixture of nitrous oxide and oxygen (7 : 3), the percentage of CO2 in the expired air being maintained between 3·5 and 4·0 % by adjustment of respiratory rate and volume of the pump. A virtual oscilloscope system (MacLab 8, Analog Digital Systems, Sydney, Australia) was used to monitor the electrocardiogram. The level of anaesthesia was monitored using electrocardiographic criteria (in particular changes in heart rate in response to noxious stimuli, as detailed in Rosa et al. 1993) and the level of spontaneous activity in the brain, captured via microelectrode penetrations at sites away from the cortical lesion (Rosa et al. 1995).
Protection of the cornea, dioptric correction and control for eye movements
Administration of atropine (1 %) and phenylephrine hydrochloride (10 %) eye drops resulted in mydriasis and cycloplegia. Application of a hard contact lens with a curvature radius of 5·25-5·4 mm focused the left eye (contralateral to the lesion) on the surface of a translucent hemispheric screen (radius, 60 cm), and protected the cornea from desiccation. Repeated ophthalmoscopic inspections revealed that optic media quality remained stable throughout the recordings. The right eye was also covered with a contact lens for protection, but was kept covered by an opaque metal shield except for occasional checks (e.g. when a cell responded poorly to the left eye, or when responses were sampled in the hemisphere contralateral to the lesion). A reversible ophthalmoscope was used to project the position of the optic discs onto a hemispheric screen. Based on our previous experience with this species (Rosa et al. 1993, 1994), the relatively high dose of pancuronium results in a stable eye position throughout the experiment, within 1-1·5 deg of the initial estimate. Indeed, analysis of receptive fields recorded at nearby sites, at different times during the experiment, confirmed the stability of the eye position. The positions of the vertical and horizontal meridians of the visual field were estimated based on the mean distance between these axes and the blind spot (Rosa et al. 1993).
Electrophysiological recordings: equipment and procedures
Tungsten-in-glass microelectrodes with an exposed tip of 10 µm were inserted in the vertical stereotaxic planes. Amplification and filtering of the electrophysiological signal was achieved via an AM Systems Model 1800 Microelectrode AC amplifier (AM Systems, Everett, WA, USA) and a 50 Hz eliminator (HumBug, Quest Scientific, Vancouver, Canada). The signal was further processed by a PC-based waveform discrimination system (SPS-8701, Signal Processing Systems, Adelaide, Australia, operated on a 133 MHz Pentium computer), which allowed the isolation of single unit spike trains with high temporal resolution. For quantitative analyses, the spike trains processed by SPS-8701 were collected via a high-fidelity interface (ITC-16 Nubus, Instrutech Corp, Great Neck, NY, USA) into a Macintosh Power PC 604e/120 MHz computer, which also controlled visual stimulus generation (see below).
In two of the experiments, where the main focus was to define the visuotopy and response strength of V2 cells, we relied mostly on the use of loudspeakers and an oscilloscope to monitor the responses. In these cases, only a few unit responses were analysed quantitatively. In one animal with a V1 lesion and one control animal, mapping with hand-held stimuli was kept to a minimum consistent with delimiting the rostral and caudal borders of V2. In these cases, the focus was on studying quantitatively the response properties of single units and small multiunit clusters. To characterize the visuotopic organization of V2, and to locate receptive fields for quantitative analysis, luminous white spots (1-10 deg in diameter) and bars (2-20 deg long, 0·2-1 deg wide) were moved on the surface of the hemispheric screen, via a hand-held projector. The bars were used when the unit was deemed to be orientation or direction biased; otherwise, the spots were preferred. To quantify the responses of V2 cells and multiunit clusters, circular patches containing high-contrast drifting sine wave gratings were generated by a stimulus and data collection software package (A/D Vance 3.55, McKellar Designs, Vancouver, BC, Canada). These stimuli were presented within the boundaries of the hand-mapped receptive fields, using an Apple Multiple Scan 20" monitor (Apple Computer Inc., Cupertino, CA, USA) located 40 cm from the eyes. The screen resolution was set to 1152 × 870 pixels, and the refresh rate to 75 Hz. Following an initial qualitative estimate of the optimal patch diameter, grating spatial frequency and drift speed, each cell was tested to characterize the optimal grating orientation and direction of motion, by varying the drift direction in 30 deg steps. Tests to define the optimal spatial frequency and drift speed were then sequentially conducted, using only the direction of motion which elicited the greatest unit response and a direction 180 deg opposite to it.
Histology
At the end of the experiment, the animal was given a lethal dose of sodium pentobarbitone (100 mg kg-1) and perfused transcardially with 0·9 % saline, followed by 4 % paraformaldehyde in 0·1 M phosphate buffer, and 4 % paraformaldehyde/10 % sucrose in phosphate buffer. Once the brain was removed from the skull, reference needles were inserted, and the block placed into fixative. The cerebellum was removed, and the block was sectioned into 40 µm slices in the parasagittal plane, using a freezing microtome. Alternate series were stained using cresyl violet and gold chloride for the visualization of myelin (three animals) or cresyl violet and cytochrome oxidase (one animal), allowing the reconstruction of electrode tracks, architectonical boundaries (using criteria defined by Rosa et al. 1994) and the extent of the cortical lesion.
Estimate of the lesion extents
The estimate of the extent of the cortex removed in each case (e.g. dark grey areas in Fig. 1) was based on a comparison between the lesioned and non-lesioned hemispheres, using photographs of the brain from dorsal, ventral and caudal perspectives. The transition between V1 and V2 was located using architectonic criteria, which included the thickness and sublaminar structure of layer 4 (Rosa et al. 1994). In one animal, in which the border region was lesioned, the rostral limit of V1 was estimated by comparison with sections from corresponding stereotaxic levels of non-lesioned hemispheres. Once the lesion extent was estimated in a series of parasagittal sections, we used receptive field maps from an extensive series of electrophysiological recordings in this species (e.g. Rosa et al. 1993) to assess the visuotopic extent of the lesion in each case. Since the projection from V1 to V2 arises from the supragranular layers (M. G. P. Rosa & A. P. Funk, unpublished results), this assessment was based on the lesion extent in the supragranular cortex.
While we acknowledge the approximate nature of the estimates thus obtained, the imprecision is unlikely to affect the conclusions of the present study. According to our previous work, both the gyral configuration and the visuotopic map of V1 are highly stereotyped in flying foxes, allowing one to use the topographic relationship between those two elements as a reliable guide. Moreover, receptive fields of V1 cells are small (1-2 deg in the central representation, and 10-15 deg in the far periphery), thus reducing the margin of error for the estimate.
|
RESULTS |
The main findings of this study are as follows. The responsiveness of V2 neurones to visual stimuli was not abolished either by partial or complete lesions of V1. In addition, the visuotopic organization of V2 in flying foxes with V1 lesions was found to be the same as that in normal controls. The data revealed a reduction in the responsiveness of cells located within the 'lesion projection zone' of V2 (i.e. the representation of the sector of the visual field corresponding to the lesioned part of V1). However, they also showed that V2 cells maintain a level of response selectivity despite the absence of V1 inputs. To substantiate these claims, we shall first illustrate the location and extent of lesions made in V1, to allow an estimate of the extent of the lesion projection zones (LPZs) in V2. We will then illustrate the visuotopy of V2 in these animals, and compare the strength of visually evoked responses inside and outside the LPZs. Finally, examples of visual responses elicited from cells within the LPZs will be illustrated.
Location and extent of lesions
The striate cortex was removed via subpial aspiration in three animals. In case AF1 (Fig. 1, top) the lesion destroyed most of area V1, while sparing the entire extent of area V2. The non-ablated portions of V1 included the region immediately adjacent to V2 (on the dorsal surface of the brain) and the dorsal bank of the splenial sulcus. Based on the comparison of parasagittal sections from the lesioned hemisphere of this animal and control data (Fig. 5 of Rosa et al. 1993), we estimate that this lesion affected the representation of parts of the visual field centred on the horizontal meridian (HM), extending to eccentricities over 50 deg, but sparing the area centralis and the vertical meridian (VM). In AF3 (not illustrated) a very similar lesion was produced, with the exception that the lesion was slightly more caudal, resulting in a more extensive sparing of the visual field around the VM.
In case AF5 (Fig. 1, bottom) the lesion was more extensive, resulting in a near complete ablation of V1, as well as destruction of the caudal part of V2. The spared portion of V1 corresponded to the representation of the peripheral vertical meridian, in the lower quadrant of the visual field. As shown in detail below, the results of recordings obtained from the remaining part of V2 in this animal were indistinguishable from those collected within the LPZs of the other two animals.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 1. Extent of cortical lesions in two animals
A-F, parasagittal sections through flying fox occipital lobe, indicating the extent of cortical lesions at different mediolateral levels. A-C are from animal AF1 (dorsal view of the brain shown in the top left panel), while D-F are from AF5 (dorsal view of the brain shown in the bottom right panel). The dorsal views indicate the extent of V1 and V2 (based on cyto-architectural criteria), the levels of the sections, and the extent of cortex ablated (grey) in each animal. In the parasagittal sections, the highly granular layer 4 (which, in flying foxes, distinguishes V1 and V2 together from surrounding areas; Rosa et al. 1994) is indicated in light grey; the borders of V1 are shown by thin dotted lines across the cortical layers, and those of V2 are shown by arrows or question marks (when the presumptive border region was included in the lesion). The estimated extent of the lesion (based on comparison with the contralateral hemisphere) is shown in dark grey and dashed outline. Abbreviations: V1, first visual area; V2, second visual area; lgn, lateral geniculate nucleus; sc, superior colliculus.
|
Visuotopic organization of V2
In order to help in the interpretation of the present data, we illustrate in Fig. 2, a summary of the normal visuotopy of V1 and V2 in Pteropus poliocephalus (based on data published by Rosa et al. 1993, 1994). As in other mammals, the representations of the area centralis and the VM form the caudal border of V2, with V1. The representation of the HM follows a curved trajectory from caudomedial to rostrolateral. The representation of the upper visual quadrant in V2 is continuous, occupying most of the lateral half of V2. In contrast, there is a field discontinuity in the representation of the peripheral lower visual quadrant in V2 (asterisks in Fig. 2). As a result of this arrangement, the portions of the lower quadrant within 20-30 deg of the VM are represented medially in V2, and those away from VM are represented laterally, rostral to the representation of the upper quadrant.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 2. Normal visuotopy and extent of V1 and V2 in the flying fox
Top left, dorsal view of the right hemisphere of a flying fox brain, showing the extent of V2 in a non-lesioned animal (data collected as part of a previous study (Rosa et al. 1994)). The region within the dashed box is magnified on the bottom left, which illustrates the visuotopy of V2 as viewed from the surface of the brain, according to the following symbols:  , representation of the area centralis;  , representation of the vertical meridian;  , representation of the horizontal meridian; thin continuous lines, representations of iso-azimuth lines; thin dashed lines, representations of iso-elevation lines. The asterisks ( ) indicate the representations of an imaginary oblique line crossing the lower visual quadrant, which corresponds to a field discontinuity in the V2 map (Rosa et al. 1994). These symbols are summarized in the chart illustrated in the upper middle panel. Bottom right: graphically 'unfolded' view of areas V1 and V2, showing the relative size and topographic relationship between these areas (symbols as above). The regions shaded in grey are normally hidden from view, being located along the medial wall, on the dorsal bank of the splenial sulcus, and on the tentorial surface. To create this representation, the map of V1 had to be artificially split along the representation of the horizontal meridian, as indicated by the arrows connecting the 60 and 90 deg iso-azimuth lines. This is necessary due to the high intrinsic curvature of the cortex containing V1 (Rosa et al. 1993).
|
Figure 3 illustrates data on the visuotopic organization of V2 in one animal with a V1 lesion. The top part of this figure shows the extent of the lesion in this animal (left), and the locations of the electrode penetrations in V2 (right). To facilitate the description, the electrode penetrations were joined in caudal-to-rostral sequences (Fig. 3A-F). The bottom six diagrams illustrate the extent of the neuronal receptive fields observed at sites corresponding to each of these sequences, as well as the estimated extent of the visual field sector represented within the lesioned part of V1 (grey). Two points should be noted. First, neuronal receptive fields included the sector of the visual field expected to be affected by the V1 lesion; in fact, some sequences (e.g. D1-9) are entirely formed by fields within this sector. Second, the visuotopy of V2 is exactly that expected on the basis of recordings from normal animals. For example, sequences of receptive fields corresponding to caudal-rostral sequences follow gradual trajectories in the visual field, moving away from the VM (e.g. Fig. 3, fields D1-9). Responses in the upper visual quadrant can be elicited from recording sites in the lateral portion of V2 (e.g. sequence E, fields 1-10). The peculiar 'split' representation of the lower quadrant is also preserved, with the region near the VM being represented medially (e.g. Fig. 3, fields A1-8) and the regions away from this meridian rostrolaterally (e.g. Fig. 3, fields E11-17). There is no evidence of a displacement of receptive fields towards the borders of the sector of the visual field affected by the V1 lesion, as has been described in monkey extrastriate cortex after V1 lesions (Rosa & Elston, 1997) or in cat V1, after retinal lesions (Kaas et al. 1990; Schmid et al. 1996).
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 3. Visuotopic organization of V2 in an animal with V1 lesion
Top left, dorsal view of the brain of animal AF1, showing the extent of V2 (grey) and of the V1 lesion (black). The region encompassed in the dotted box is magnified in the top right panel, which also illustrates recording sites obtained in V2, joined to form six anteroposterior sequences (A-F). The recording sites in each sequence are numbered sequentially, from caudal (site 1) to rostral. The interpolated representation of the horizontal meridian in V2 is indicated by a dotted line. The bottom 6 diagrams (A-F) are charts of the contralateral hemifield, indicating the estimate of the region of the visual field that was represented in the lesioned sector of V1 (grey), the blind spot (black), and the receptive fields of neurones recorded at each site in V2. The borders of receptive field E 17 extended beyond the 90 deg azimuth, as indicated by the arrows. Abbreviations: HM, horizontal meridian; VM, vertical meridian.
|
Examples of data collected in case AF5 are shown in Fig. 4. In spite of the fact that part of V2 was ablated in this animal, the evidence suggests that the visuotopy of V2 has been preserved, despite a near complete ablation of V1. Sequences of electrode penetrations near the rostrolateral border of V2 (Fig. 4, top right) resulted, as expected, in receptive fields located in the visual field periphery (50-80 deg). Moreover, the expected displacement of receptive fields towards the periphery was detected, as penetrations were placed successively further from the V1-V2 border (compare, for example, sequences A and C in Fig. 4). Finally, in each of the sequences, the receptive fields moved towards the upper visual quadrant, as the recording sites became more lateral (cf. fields 1, 6 and 12 in Fig. 4, sequence A).
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 4. Receptive fields of V2 cells in an animal with an extensive V1 lesion
Top left, brain of animal AF5, showing the extent of V1, V2 and a cortical lesion (grey). A magnified view of the brain, based on a photograph taken from a perspective normal to the surface of V2, is illustrated top right. This panel also illustrates the locations of three rows of recording sites crossing the peripheral representation of V2 (A-C), from the lower quadrant representation (-) to the upper quadrant representation (+). The bottom 3 diagrams (A-C) are charts of the contralateral hemifield, indicating the estimate of the region of the visual field that was represented in the lesioned sector of V1 (grey) and the receptive fields of neurones recorded in sites corresponding to rows A-C. Note the peripheral displacement of receptive field sequences from A to C, and the change of receptive field positions from the first to the last site of each row.
|
Neuronal responses within and outside the LPZs
Figure 5 compares the overall quality of responses obtained from cells in different parts of V2 in three lesioned animals and one non-lesioned control. This comparison is based on a subjective ranking, made on the basis of examination of oscilloscope traces and monitoring of responses via loudspeakers during the experiment. Responses were deemed 'strong' if they represented a clear increase relative to the spontaneous activity and if they were consistent between trials (i.e. the neurone increased its firing rate during every presentation of the stimulus). The 'weak' category included cells that rapidly habituated to the presentation of the stimulus, typically requiring long intertrial intervals (10 s or more) to yield responses, or that yielded few spikes in response to stimulation. 'Non-responsive' cells were those which showed spontaneous firing, but did not respond to any of the stimuli presented.
Analysis of Fig. 5 shows that the responsiveness of V2 cells outside the LPZ in lesioned animals was only slightly lower than that observed in the non-lesioned control animal. The percentage of non-LPZ cells with strong responses (72·3 %) is close to that found in the control recordings throughout V2 (70·8 %); however, lesioned animals showed a higher proportion of non-responsive neurones than the control (17·0 vs. 4·2 %). These two distributions are significantly different (
2 = 50·1, P < 0·01, d.f. = 2).
The cellular responses within the LPZs were depressed in comparison with either non-LPZ or control cells. As illustrated in Fig. 5, only 31·5 % of the V2 neurones with receptive fields located within the sector of the visual field corresponding to the lesioned part of V1 yielded strong responses, while 40·2 % did not respond to any of the stimuli we employed. These proportions are very different from those observed amongst cells outside the LPZ (2 = 79·6, P < 0·01, d.f. = 2). Figure 5 also shows that the level of responsiveness was similar whether one considers only those V2 cells with receptive fields partially overlapping the estimates of the sector of the visual field affected by the V1 lesion, or those with receptive fields completely within this region (2 = 2·9, P = 0·23, d.f. = 2). In summary, although the majority of neurones in V2 remained visually responsive following a lesion of V1, there was a significant decline in the quality of these responses, and an increase in the proportion of non-responsive cells.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 5. Responsiveness of neurones in different portions of V2
Histograms illustrating the percentages of neurones showing strong visual responses ( ), weak visual responses () and no detectable visual response ( ) in different parts of V2 of 3 lesioned animals, as compared with a control animal studied with identical techniques (Non-lesioned). The number of cells contributing to the sample is shown above each histogram. Abbreviation: LPZ, lesion projection zone.
|
We have examined in detail the response properties of seventeen cells in V2 of lesioned animals. Of those, only four proved to have receptive fields entirely restricted to the sector of the visual field expected to be affected by the V1 lesion, on the basis of histological reconstruction. While this sample is too small to allow a statistically reliable comparison with V2 cells in the normal animal (n = 7), it confirms our assessment, based on qualitative techniques, that many V2 cells remain visually responsive after removal of V1. Figure 6 illustrates an example of a LPZ cell which was qualitatively deemed to have a 'strong' response. Like all LPZ cells we had the opportunity to study, this neurone responded best to stimuli at low spatial and temporal frequencies. In this particular case, a sustained response required stimulation with a pattern of 0·025 cycles deg-1, at a temporal frequency of 1 Hz (Fig. 6B, top histogram). However, brief, transient responses could be elicited by stimuli at spatial frequencies of up to 0·125 cycles deg-1 (Fig. 6B, bottom histogram) and temporal frequencies of over 5 Hz (Fig. 6C). We failed to observe sharp orientation and direction selectivity amongst the cells studied in the LPZ, a result that may be a function of the spatiotemporal parameters of the 'optimal' stimuli. Examples of cells with slight but significant direction and orientation biases are illustrated in Figs 6 and 7. Typically for LPZ cells, the response of the neurone shown in Fig. 7 tends to habituate rapidly in such a way that the activity drops dramatically after 1-2 s of repeated stimulation, even when near-optimal spatiotemporal parameters are used.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 6. Responses of a V2 neurone located within the LPZ
A, polar plot showing the responses of a V2 layer 4 neurone with a receptive field located entirely within the sector of the visual field corresponding to the V1 lesion. The mean response rate was calculated within a 1 s window starting with the presentation of a stimulus (in this case, a sine wave grating pattern, 0·04 cycles deg-1, presented at a temporal frequency of 1·5 Hz) drifting in 12 directions. This particular cell had no spontaneous activity and showed a slight bias for movement towards the lower temporal quadrant. This direction was used for subsequent tests (B and C). B, effect of varying the spatial frequency of the stimulus, whilst keeping the orientation and drift speed constant (40 deg s-1). Grating patterns of variable spatial frequency were presented for 3 s (each histogram corresponds to the accumulation of 10 randomly interleaved trials). A sustained response could only be obtained for a very low spatial frequency pattern (top histogram). C, effect of varying the stimulus drift speed, whilst maintaining the orientation and spatial frequency constant (0·025 cycles deg-1). Sustained responses depended on low temporal frequencies (0·67-1·8 Hz; see top 3 histograms).
|
Figure 8 illustrates the responses of a V2 neurone with a receptive field located in a region of the visual field that, according to our estimate, was represented in the portion of V1 spared by the lesion in animal AF3. Although the number of cells studied is admittedly small, we observed that non-LPZ neurones responded with higher peak firing rates, and were sensitive to higher spatial frequencies, than LPZ cells. Other response properties did not obviously differ between cells within and outside the V2 LPZs. Thus, similar to the data illustrated in Figs 6 and 7, the neurone illustrated in Fig. 8 had a broad direction selectivity (Fig. 8A), and responded best to a relatively slow drifting grating pattern (18-36 deg s-1; Fig. 8B). However, unlike cells located within the LPZs at a comparable eccentricity (20 deg), this neurone responded in a sustained manner to spatial frequencies of up to 0·15 cycles deg-1 (Fig. 8C), corresponding to temporal frequencies over 5 Hz. Neurones studied in the control animal (AF4) had similar response properties.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 7. Responses of a V2 neurone located within the LPZ
A and B, responses of a V2 neurone with a receptive field located entirely within the sector of the visual field corresponding to the V1 lesion. This cell was located near the limit between layers 4 and 5. Each histogram is based on the accumulation of 10 randomly interleaved trials. A, effect of varying the stimulus drift speed, whilst maintaining the spatial frequency near that which was optimal for the cell (0·05 cycles deg-1). This cell showed habituation during the stimulus presentation (3 s), even for the lowest drift speed tested (18 deg s-1, corresponding to a temporal frequency of 0·9 Hz). The high temporal frequency cut-off for the transient response was just over 6 Hz (bottom two histograms). B, neuronal response bias for grating orientation. The histograms illustrated were accumulated during the presentation of gratings (0·05 cycles deg-1, 27 deg s-1) drifting in each of 12 directions. In each trial, grating orientation was perpendicular to the axis of motion.
|
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 8. Responses of a V2 neurone located outside the LPZ
A-C, responses of a V2 neurone with a receptive field located entirely outside the sector of the visual field corresponding to the V1 lesion. This cell was located in lower layer 3. Each histogram is based on the accumulation of 10 randomly interleaved trials. A, polar plot showing the responses of the neurone to a sine wave grating pattern (0·1 cycles deg-1, presented at a temporal frequency of 3·6 Hz) drifting in 12 directions. This cell showed a broad selectivity for movement towards the lower nasal quadrant. B, effect of varying the stimulus drift speed, whilst maintaining the spatial frequency (0·125 cycles deg-1) constant. The first histogram in this series illustrates the response of the neurone to a stationary grating, presented on the screen for 2 s. C, effect of varying the spatial frequency of the stimulus, whilst maintaining a constant orientation and drift (36 deg s-1). A sustained response could be obtained for spatial frequencies up to 0·15 cycles deg-1.
|
|
DISCUSSION |
The main findings of this study are that partial or complete lesions of V1 are not sufficient to suppress visual responsiveness of a large proportion of V2 neurones, and that the normal visuotopic organization of V2 is retained following a lesion of V1. These results demonstrate that the functional relationship of V2 to area V1 in flying foxes is unlike that found in primates.
Serial versus parallel processing in visual cortex
Explanations of the mechanisms behind mammalian vision, and specifically of the interrelationships between cortical areas, have been a persisting source of debate. For example, the interpretation of early studies in the cat suggested a serial processing scheme, whereby a visual stimulus would sequentially activate cells in areas V1 (area 17), V2 (area 18) and V3 (area 19), resulting in the 'construction' of receptive fields with gradually more complex stimulus requirements (e.g. Hubel & Wiesel, 1965). However, later work has forced a revision of this scheme, in so far as areas V1 and V2 are concerned. The functional characteristics of the thalamic inputs to V2 cells are different from those of V1 cells, and the response properties of cells in these areas are incompatible with the notion that the response selectivities of V2 cells are built from V1 inputs (e.g. Stone & Dreher, 1973; Tretter et al. 1975; Pettigrew & Dreher, 1987). Moreover, the responsiveness of V2 cells survives lesion or inactivation of V1 (e.g. Dreher & Cottee, 1975; Sherk, 1978; Casanova et al. 1992), and the distributions of latencies of neuronal responses in these areas overlap extensively (the latencies of V2 cells being, if anything, lower than those of V1 cells; Dinse & Krüger, 1994). Studies in which the striate cortex of rats was removed have also demonstrated a normal visuotopic organization in lateral prestriate cortex, including the lateromedial area, LM, which is the putative homologue of V2 (Olavarria & Torrealba, 1978). However, in both cat and rat, an overall reduction of visual responsiveness in V2 was apparent following inactivation or removal of area V1 (Sherk, 1978; Molotchnikoff & Hubert, 1990), as were modifications in the velocity, direction and/or orientation selectivity of neurones in area V2 (Donaldson & Nash, 1975; Dreher & Cottee, 1975; Sherk, 1978; Casanova et al. 1992; Chabli et al. 1998). One of the main questions left open by the present study is the extent to which V2 responses are changed by V1 inactivation. While our data do demonstrate that some V2 cells respond in an orientation- and direction-biased manner after V1 lesions, it is conceivable that the proportion of such selective cells, and/or the tuning bandwidth of specific cells, do change. Studies of a larger number of neurones, preferably employing reversible inactivation of neuronal populations in V1, would be necessary to evaluate these possibilities.
Research on simian primates has found no evidence of residual activity in V2 after lesion or inactivation of V1 (Cowey, 1964; Schiller & Malpeli, 1977; Girard & Bullier, 1989; M. G. P. Rosa & G. N. Elston, unpublished results). Thus, V2 function in primates seems to be critically dependent on the integrity of V1. Whether this observation implies a strictly serial processing (implying that the responses of V2 cells are 'built' primarily through the combination of inputs from V1 cells) is a matter of contention. Although some models of visual processing in the monkey have emphasized the hierarchical alternative (e.g. Felleman & Van Essen, 1991), others have placed emphasis on the separate pathways that convey thalamic innervation to V1 and V2 (e.g. Creutzfeldt, 1988). Indeed, as in the cat, the response latencies of cells in primate V1 and V2 overlap extensively, and cells in these areas may fire synchronously (Frien et al. 1994; Bullier & Nowak, 1995). While it is probable that the processing of information in simian areas V1 and V2 is, to some extent, serial in nature, it should be kept in mind that the fact that V2 cells do not respond in the absence of V1 could, in theory, also be accounted for by non-serial mechanisms. For example, V1 could provide a critical non-specific tonic activating input, without which the thalamic activation would be insufficient to bring V2 cells to firing threshold.
Why are primates different?
At present, there are limited data on which to base discussion of which, if any, anatomical characteristics make simian primates different from cats, rats and flying foxes, with regards to the dependence of V2 on its V1 inputs. However, one can identify at least two factors that may be important in this context: the existence of a direct projection from the main layers of the LGNd to V2, and the relative development of the retinogeniculate and retinotectothalamic pathways in a given species. Although it is true that the projection from the LGNd to V2 is minor in the monkey as compared with the cat (Bullier & Kennedy, 1983), it is unlikely that robust V2 responses after inactivation of V1 are exclusively a function of the strength of the geniculoprestriate projection in a given species. For example, rodents seem to lack a significant geniculoprestriate projection (Ribak & Peters, 1975; Olavarria, 1979), in spite of the fact that V2 responsiveness is relatively independent of V1 inputs. Indeed, the lack of heavy LGNd projections to V2 seems to be a relatively widespread characteristic among mammals, rather than a primate attribute.
Among animals that lack a strong geniculoprestriate projection, the ability of V2 cells to respond after inactivation of V1 could depend on the relative dominance of the visual system by the retinogeniculostriate pathway. Perhaps the fact that V2 in monkeys becomes unresponsive after V1 lesion is related to the enormous expansion of this system in the lineage leading to present-day primates; for example, the percentage of retinal ganglion cells that project to the LGNd is much higher in the monkey (90 %) than in the rat (35 %; Garey et al. 1991). It is conceivable that this expansion, which has been linked to the development of a fine acuity system (Rosa et al. 1997), has led to a greater dependence of V2 on the information channelled through V1. By contrast, in the rat, where most retinal axons project to the superior colliculus, the tectothalamicprestriate projections are sufficient to convey visual information to V2.
While the anatomical relationships of thalamic nuclei with V1 and V2 in the flying fox are still under investigation, there is evidence of a projection from the main LGNd layers to V2 (M. G. P. Rosa & A. P. Funk, unpublished observations). In addition, we have observed that in the flying fox, as in the cat (and unlike in rodents and primates), layer 4 in both areas V1 and V2 stains heavily with cytochrome oxidase (Rosa et al. 1994), a characteristic that has been correlated with the sites of termination of LGNd projections (Wong-Riley, 1979; Price, 1985). It would seem that strong projections from the main LGNd layers to prestriate cortex (including V2) have appeared independently in different mammalian lineages (including carnivores and megachiropterans), rather than being a remnant of a primitive situation in which both the LGNd and the lateral posterior complex projected extensively throughout visual cortex (Diamond & Hall, 1969).
Implications for the phylogeny of flying foxes
The phylogeny of the flying fox has been surrounded by controversy since Pettigrew (1986) made the surprising finding that flying foxes have an advanced organization of the retinotectal pathway, previously thought to be exclusive to primates (for neurophysiological evidence against and in support of these findings see Thiele et al. 1991 and Rosa & Schmid, 1994, respectively). This observation, as well as other neuroanatomical characteristics (including the organization of the LGNd and the cellular composition of the hippocampus; Pettigrew et al. 1989) led to the proposal that flying foxes and dermopterans (gliding lemurs) are extant representatives of groups that diverged from the same early mammalian stock as primates, with the implication that flight evolved independently in two mammalian groups: microchiropteran bats and flying foxes.
Our results do not reflect those expected if the visual pathways of flying foxes were organized similarly to those of simian primates. Instead, they suggest that the functional relationship between V1 and V2 in the flying fox is likely to represent a condition that is widespread among extant mammalian orders. Thus, while not excluding the possibility that flying foxes do form a sister group of primates, our data suggest that the dependence of V2 activity on V1 inputs has emerged only in simian primates. In this context, it will be particularly important to establish whether or not V2 depends upon inputs from V1 in pro-simian primates, animals in which V1 and V2 present a series of characteristics intermediate between those found in non-primate mammals and simian primates (Rosa et al. 1997), and other candidate primate sister groups, such as tree shrews (Jain et al. 1994). Although the relevant physiological studies have not been performed yet, behavioural tests have shown that lesions of V1 in the pro-simian Galago result in a sensory loss similar to that observed in monkeys, while similar lesions in tree shrews cause far less severe visual deficits (Atencio et al. 1975).
|
REFERENCES |
| Atencio, F. W., Diamond, I. T. & Ward, J. P. (1975). Behavioral study of the visual cortex of Galago senegalensis. Journal of Comparative and Physiological Psychology 89, 1109-1135. |
[Medline] |
| Bullier, J. & Kennedy, H. (1983). Projection of the lateral geniculate nucleus onto cortical area V2 in the macaque monkey. Experimental Brain Research 53, 168-172 |
[Medline] |
| Bullier, J. & Nowak, L. G. (1995). Parallel versus serial processing: new vistas on the distributed organization of the visual system. Current Opinion in Neurobiology 5, 497-503 |
[Medline] |
| Casanova, C., Michaud, Y., Morin, C., McKinley, P. A. & Molotchnikoff, S. (1992). Visual responsiveness and direction selectivity of cells in area 18 during local reversible inactivation of area 17 in cats. Visual Neuroscience 9, 581-593 |
[Medline] |
| Chabli, A., Ruan, D. Y. & Molotchnikoff, S. (1998). Influences of area 17 on neuronal activity of simple and complex cells of area 18 in cats. Neuroscience 84, 685-698 |
[Medline] |
| Cowey, A. (1964). Projection of the retina on to striate and prestriate cortex in the squirrel monkey, Saimiri sciureus. Journal of Neurophysiology 27, 366-393. |
|
| Creutzfeldt, O. D. (1988). Extrageniculostriate visual mechanisms: compartmentalization of visual functions. Progress in Brain Research 75, 307-320 |
[Medline] |
| Diamond, I. T. & Hall, W. C. (1969). Evolution of neocortex. Science 164, 251-262 |
[Medline] |
| Dinse, H. R. & Krüger, K. (1994). The timing of processing along the visual pathway in the cat. NeuroReport 5, 893-897 |
[Medline] |
| Donaldson, I. M. & Nash, J. R. (1975). The effect of a chronic lesion in cortical area 17 on the visual responses of units in area 18 of the cat. The Journal of Physiology 245, 325-332 |
[Abstract] |
| Dreher, B. & Cottee, L. J. (1975). Visual receptive field properties of cells in area 18 of the cat's cerebral cortex before and after acute lesions in area 17. Journal of Neurophysiology 38, 735-750 |
[Medline] |
| Felleman, D. J. & Van Essen, D. C. (1991). Distributed hierarchical processing in the primate cerebral cortex. Cerebral Cortex 1, 1-47 |
[Abstract] |
| Frien, A., Eckorn, R., Bauer, R., Woelbern, T. & Kehr, H. (1994). Stimulus-specific fast oscillations at zero phase between visual areas V1 and V2 of awake monkey. NeuroReport 5, 2273-2277 |
[Medline] |
| Garey, L. J., Dreher, B. & Robinson, S. R. (1991). The organization of the visual thalamus, In Vision and Visual Dysfunction vol. 3, Neuroanatomy of the Visual Pathways and their Development, ed. Dreher, B. & Robinson, S. R., pp. 176-234. Macmillan, London. |
|
| Girard, P. & Bullier, J. (1989). Visual activity in area V2 during reversible inactivation of area 17 in the macaque monkey. Journal of Neurophysiology 62, 1287-1302 |
[Medline] |
| Hubel, D. H. & Wiesel, T. N. (1965). Receptive fields and functional architecture in two non-striate visual areas (18 and 19) of the cat. Journal of Neurophysiology 28, 229-289. |
|
| Jain, N., Preuss, T. M. & Kaas, J. H. (1994). Subdivisions of the visual system labeled with the Cat-301 antibody in tree shrews. Visual Neuroscience 11, 731-741 |
[Medline] |
| Kaas, J. H., Krubitzer, L. A., Chino, Y. M., Langston, A. L., Polley, E. H. & Blair, N. (1990). Reorganization of retinotopic cortical maps in adult mammals after lesions of the retina. Science 248, 229-231 |
[Medline] |
| Molotchnikoff, S. & Hubert, F. (1990). Susceptibility of neurons in area 18a to blockade of area 17 in rats. Brain Research 510, 223-228 |
[Medline] |
| Olavarria, J. (1979). A horseradish peroxidase study of the projections from the latero-posterior nucleus to three lateral peristriate areas in the rat. Brain Research 173, 137-141 |
[Medline] |
| Olavarria, J. & Torrealba, F. (1978). The effect of acute lesions of the striate cortex on the retinotopic organization of the lateral peristriate cortex in the rat. Brain Research 151, 386-391 |
[Medline] |
| Pettigrew, J. D. (1986). Flying primates? Megabats have the advanced pathway from eye to midbrain. Science 231, 1304-1306 |
[Medline] |
| Pettigrew, J. D. & Dreher, B. (1987). Parallel processing of binocular disparity in the cat's retinogeniculocortical pathways. Proceedings of the Royal Society B 232, 297-321 |
[Medline] |
| Pettigrew, J. D., Jamieson, B. G., Robson, S. K., Hall, L. S., McAnally, K. I. & Cooper, H. M. (1989). Phylogenetic relations between microbats, megabats and primates (Mammalia: Chiroptera and Primates). Philosophical Transactions of the Royal Society B 325, 489-559. |
|
| Price, D. J. (1985). Patterns of cytochrome oxidase activity in areas 17, 18 and 19 of the visual cortex of cats and kittens. Experimental Brain Research 58, 125-133 |
[Medline] |
| Ribak, C. E. & Peters, A. (1975). An autoradiographic study of the projections from the lateral geniculate body of the rat. Brain Research 92, 341-368 |
[Medline] |
| Rosa, M. G. P., Casagrande, V. A., Preuss, T. & Kaas, J. H. (1997). Visual field representation in striate and prestriate cortices of a prosimian primate (Galago garnetti). Journal of Neurophysiology 77, 3193-3217 |
[Abstract/Full Text] |
| Rosa, M. G. P. & Elston, G. N. (1997). Responsiveness of neurones in visual area MT after striate cortex lesions. Proceedings of the Australian Neuroscience Society 8, 63. |
|
| Rosa, M. G. P. & Schmid, L. M. (1994). Topography and extent of visual-field representation in the superior colliculus of the megachiropteran Pteropus. Visual Neuroscience 11, 1037-1057 |
[Medline] |
| Rosa, M. G. P., Schmid, L. M. & Calford, M. B. (1995). Responsiveness of cat area 17 after monocular inactivation: limitation of topographic plasticity in adult cortex. The Journal of Physiology 482, 589-608 |
[Abstract] |
| Rosa, M. G. P., Schmid, L. M., Krubitzer, L. A. & Pettigrew, J. D. (1993). Retinotopic organization of the primary visual cortex of flying foxes (Pteropus poliocephalus and Pteropus scapulatus). Journal of Comparative Neurology 335, 55-72 |
[Medline] |
| Rosa, M. G. P., Schmid, L. M. & Pettigrew, J. D. (1994). Organization of the second visual area in the megachiropteran bat Pteropus. Cerebral Cortex 4, 52-68 |
[Abstract] |
| Sandell, J. H. & Schiller, P. H. (1982). Effect of cooling area 18 on striate cortex cells in the squirrel monkey. Journal of Neurophysiology 48, 38-48 |
[Medline] |
| Schiller, P. H. & Malpeli, J. G. (1977). The effect of striate cortex cooling on area 18 cells in the monkey. Brain Research 126, 366-369 |
[Medline] |
| Schmid, L. M., Rosa, M. G. P., Calford, M. B. & Ambler, J. S. (1996). Visuotopic reorganization in the primary visual cortex of adult cats following monocular and binocular retinal lesions. Cerebral Cortex 6, 388-405 |
[Abstract] |
| Sherk, H. (1978). Area 18 cell responses in cat during reversible inactivation of area 17. Journal of Neurophysiology 41, 204-215 |
[Medline] |
| Stone, J. & Dreher, B. (1973). Projection of X- and Y-cells of the cat's lateral geniculate nucleus to areas 17 and 18 of visual cortex. Journal of Neurophysiology 36, 551-567 |
[Medline] |
| Thiele, A., Vogelsang, M. & Hoffmann, K. P. (1991). Pattern of retinotectal projection in the megachiropteran bat Rousettus aegyptiacus. Journal of Comparative Neurology 314, 671-683 |
[Medline] |
| Tretter, F., Cynader, M. & Singer, W. (1975). Cat parastriate cortex: a primary or secondary visual area. Journal of Neurophysiology 38, 1099-1113 |
[Medline] |
| Wong-Riley, M. (1979). Changes in the visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry. Brain Research 171, 11-28 |
[Medline] |
Acknowledgements
The authors would like to thank Rowan Tweedale for many comments that improved the manuscript, and Rita Collins for the superb histological work. This study was supported by grant no. 961144 of the National Health and Medical Research Council.
Corresponding author
M. G. P. Rosa: Vision, Touch & Hearing Research Centre, Department of Physiology and Pharmacology, The University of Queensland, QLD 4072, Australia.
Email: M.Rosa{at}vthrc.uq.edu.au
This article has been cited by other articles:
|
|
|
|
 
C. E. Collins, D. C. Lyon, and J. H. Kaas
Responses of Neurons in the Middle Temporal Visual Area After Long-Standing Lesions of the Primary Visual Cortex in Adult New World Monkeys
J. Neurosci.,
March 15, 2003;
23(6):
2251 - 2264.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Introduction
