Visual Neuroscience (1990), 5 , 83-98. Printed in the USA. Copyright © 1990 Cambridge University Press 0952-5238/90 $5.00 + .00
Properties of area 17/18 border neurons contributing to the visual transcallosal pathway in the cat
M.E. M c C O U R T , J. T H A L L U R I , AND G . H . H E N R Y Division of Neuroscience, John Curtin School of Medical Research, Australian National University, Canberra, Australia (RECEIVED November 20, 1989; ACCEPTED April 4, 1990)
Abstract In a series of physiological experiments, a total of 203 neurons at the Area 17/18 border were recorded with a callosal link either demonstrated by antidromic or transsynaptic activation from stimulating electrodes located in the homotopic contralateral hemisphere (CH), or in the splenial segment of the corpus callosum (CC). Forty-four percent of the transcallosal cells could also be driven from stimulating electrodes in or just above the lateral geniculate nucleus (OR,). The majority (69%) of transcallosal neurons were classifiable as belonging to the complex family (B and C cells) and most of these were found in the supragranular laminae and in lamina 4A. The ocular dominance distribution of transcallosal cells was trimodal, consisting of roughly equal numbers of monocularly dominated and binocularly balanced neurons. Estimates of conduction time and synaptic delay were obtained for neurons driven from CH, CC, and from OR], and in most instances the response latency was short enough to suggest a monosynaptic input from either the ipsi- or contra-lateral hemisphere. The distribution of transcallosal conduction times showed that S cells, as a class, had significantly faster conduction than cells of the complex family but otherwise there was no obvious signs of multimodality in the distribution curve An analysis of the synaptic delays in transcallosal activation produced a mean of 0.6 to 0.7 ms but some were too short to be consistent with a transsynaptic drive suggesting that some cells with an antidromic drive may have been included in the transsynaptic category Results are interpreted in terms of the contribution made by the corpus callosum to stereoscopic vision. Keywords: Visual cortex, Transcallosal pathways, Receptive fields, Response latency, Cat, Electrical stimulation
Introduction
calls for a revision of the contribution of transcallosal neu rons to coarse stereopsis; whereas they could act alone in the processing of uncrossed-disparity information, they may only supplement the chiasmal contribution in the processing of crossed-disparity signals. The earliest attempts to separate the contribution from the chiasm and the corpus callosum were behavioral experiments which showed that cortical binocularity and stereoscopic depth discrimination in the cat were achieved through both chiasmal and transcallosal pathways (Fox & Blake, 1971; Packwood & Gordon, 1975; Kaye et al., 1981, 1982; Timney et al., 1985; Ptito et al., 1986; Lepore et al., 1986). Several of these studies suggested that the naso-temporal overlap in the chiasm makes a major, although not exclusive, contribution to stereopsis in the cat (Lepore et al., 1986; Ptito et al., 1986). A weak contri bution also comes through the transcallosal pathway so that cats with transsectioned chiasms performed at greater-thanchance levels on disparity-discrimination tasks. The relative weakness of the callosal contribution revealed in these experi ments may provide another indication that crossed-disparity in formation exclusively used in the targets of these studies is Drocessed principally' bv BT or Y retinal Eanslion cells of the naso-temporal overlap
The decussation of the projections arising from each eye results in separate cerebral representations of the two halves of the visual field which are believed to be reunified, in part, through the corpus callosum. In an assessment of this reunion, Bishop and Henry (1971) proposed that the transcallosal linkages were only responsible for coarse stereopsis involving large disparities and that fine stereopsis was mediated through the nasotemporal overlap in the optic chiasm. Subsequent analysis of streaming in the optic chiasm prompted Levick (1977) to sug gest that the partial decussation of the brisk-transient (BT) pop ulation of cat retinal ganglion cells (equivalent to Y-cells in the classification of Enroth-Cugell & Robson, 1966) was well-suited to support crossed-disparity, coarse stereopsis. This proposal
Reprint requests to: G.H. Henry, Centre for Visual Sciences, JCSMR, P.O. Box 334, Canberra, ACT, 2601, Australia. Present address of M.E. McCourt: Department of Psychology and Institute for Neurological Sciences Research, University of Texas (Aus tin), Austin, TX, 78712, USA. Present address of J. Thalluri, School of Applied Science, South Australian College of Advanced Education, Salisbury East, South Aus tralia, 5109, Australia.
83
M.E. McCourt, J. Thalluri, and G.H. Henry
84 In a more direct approach, a variety of experimental proce dures have been employed to assess the physiological response properties of visual neurons contributing to or receiving from the transcallosal pathway. A common technique is to compare neurons in the callosal terminal zone (at the border of areas 17/18) before and after removing the input of the splenial seg ment of the corpus callosum (Payne et al., 1980; Payne et al., 1984; Minciacchi & Antonini, 1984; Cynader et al., 1986; but see Elberger & Smith, 1985, and Payne, 1986 for discussion) or before and after the midline sectioning of the corpus callosum. Both of these intrusions lead to a reduction in the proportion of binocularly activated neurons when compared to the intact cat After splitting the optic chiasm the majority of remaining neurons (viz callosally recipient) which belong to the complex cell familv (B and C cells in the terminoloev of Henrv 1977) possess characteristically laree receDtive fields and are concen irated in the supragranular layers and in upper lamina. 4 (Lepore & Guillemot 1982) A weakness in the ablation and sectioning experiments is the uncertainty that the recorded cells have a link with the corpus callosum. This deficiency is overcome in experiments where the responses of callosal fibers are recorded directly (Hubel & Wiesel, 1967; Berlucchi et al., 1967; Shatz, 1977; Berardi et al., 1987) or where responses to electrical stimulation are used to identify a callosal link (Harvey, 1980; Innocenti, 1980; McCourt et al., 1985, 1987). The direct recordings from fibers of the cor pus callosum reveals that contributing neurons may belong to any response class, that most are binocular and that many possess high-contrast thresholds and low spatial and temporal acuities more typical of cells of the Y-than the X-stream. It is possible however that these experiments using recordings from transcallosal fibers may be contaminated by fibers passing to other visual areas such as the suprasylvian complex (Segraves & Rosenauist 1982a b Rosenauist 1985" Rauschecker et al 1987) where contributing cells are known to have lower visual acuity (DiStephano et aL 1985) m
The application of electrical stimulation to the transcallosal pathway while recording from cells at the border of areas 17/18, the method employed in the present experiments, makes it possible to identify cells projecting to or receiving from the corpus callosum. For this experimental arrangement, neurons contributing to the corpus callosum respond with an antidro mic spike while those receiving a callosal input are driven transsynaptically. Experiments of this type (Harvey, 1980; Innocenti, 1980; McCourt et al., 1985, 1987) have revealed that neurons contributing to and receiving from the corpus callosum belong principally to the complex family (B or C cells), are distributed around the border of laminae 3 and 4, and are predominantly binocular in their responses The present study utilizes the methods of extracellular recording and electrical stimulation to identify callosally projecting and recipient neurons at the border of areas 17/18. The results, therefore, expand and elaborate on the earlier stud ies and more extensive sampling makes it possible to develop more decisive comparisons in the response properties of neurons contributing to and receiving from the corpus callosum. In ad dition, the larger sample of conduction velocities, assessed from response latencies to electrical stimulation at two points in the transcallosal pathway, provides a clearer picture of the time for spike initiation and synaptic delay in the corpus callosum. In a new approach, the present study also compares conduction properties of axons of the transcallosal pathway against those
in afferent fibers coming from the lateral geniculate nucleus (LGN) to the same contributing cell. Another comparison looks at the laminar distributions for cells labeled retrogradely with the axonal tracer, wheat germ agglutinin-horseradish peroxidase ( W G A - H R P ) , and relates it to the distributions of cells con tributing to and receiving from the corpus callosum. The results of the experiments are then interpreted in terms of the contri bution made to the processing of information on coarse stere opsis by signals passing through the transcallosal pathway.
Methods Subjects Twelve adult cats were studied; six were used in the electrophys iological experiments, and six for the tracing of anatomical pathways. The animals ranged from 2.5-3.5 kg in weight. All cats were healthy, possessed normal interocular alignment, and displayed no visual defects.
Electrophysiological
experiments
Preparation Surgery was performed under halothane anesthesia (4% in duction; 1% maintenance dosage) in a mixture of N 0 : 0 (70:30). A long-lasting local anesthetic (2% marcaine) was in jected around the margins of all surgical incisions. The admin istration of halothane was discontinued following surgery, and was replaced during recording sessions by the nitrous oxide to oxygen (70:30) mixture, supplemented by continuous intrave nous infusion of pentobarbitone sodium (1.0 mg k g " h " Nembutal, Bomac Laboratories, Australia; Hammond, 1978). The N O : 0 mixture, delivered through a tracheal cannula, was adjusted around a stroke volume of 40 ml, and a stroke rate of 20 m i n , to achieve a constant end-tidal C 0 level of 4 % . Temperature was measured by a rectal thermoprobe, and was held constant at 38°C with a thermostatically controlled heating blanket. Heart rate, electrocardiogram, and the electro encephalogram were monitored throughout. Paralysis of the ex traocular muscles was achieved by continuous intravenous infusion of pancuronium bromide (1.4 mg k g h Pavulon, N.V. Organon Oss, Holland) and gallamine triethiodide (5 mg kg h Flaxedil, May & Baker, Australia) in depolyte (Intervet Australia Pty Ltd ) A bilateral cervical sympathectomy was performed to further reduce the occurrence of residual eye movements Contact lenses of zero power were installed to pro tect the corneas and the evelids and nictitating membranes were retracted bv 'the tonical aoDlication of phenylephedrine HC1 (10% neosvnDhrine). The nuoils were dilated and accommoda 2
2
1
2
1
2
- 1
2
- 1
- 1
_ 1
_ 1
tinn was rplaxpH hv thp tnniral annliratinn of 1 QTh atronine sul n h » J
t
and T
H o n ih
rnrn d i 7 T e r m
o
,
?
a
T
a
\
r t , f L ^ n.mik h
T
i
of :
w e r e fitted
7e7« f„„H
and
. r a m
centered on tne optic axis wun tne a m ot a zxiss tunaus cam era, wntcn was also used to DacK-project retinal « l°P" tic disc, area centralis) onto a tangent screen at irequem intervals, ine animal was retinoscopicatly retracted and correct ing lenses were fitted accordingly. Small craniotomies were performed for the insertion of the recording and stimulating electrodes. Single units were isolated and their activity was recorded with extracellular glass-coated tungsten microelectrodes (Merrill & Ainsworth, 1972), advanced i a n a m a r K
85
Properties of area 17/18 border neurons in steps of 1 m by a hydraulic microdrive. Stimulating elec trodes were also of tungsten-in-glass construction, but has ex posed tips of approximately 1 mm. Stimulating electrodes were positioned as follows: one in the visual cortex of the contralat eral hemisphere (CH), near the area 17/18 border, at HorsleyClarke coordinates. AP:-3.0, ML: 1.5; a second in the splenium of the corpus callosum (CC) at Horsley-Clarke coordinates AP:3.0 and 5.0, ML:0; and one located in or just above the ip silateral LGN (OR!) at Horsley-Clarke coordinates AP.6.8, ML:9.0. The final position for the stimulating electrode at CH was determined by first probing with a multiunit recording elec trode in CH until units were encountered in retinotopic corre spondence with those isolated by the recording electrode in the iDsilateral cortex The CC and OR stimulating electrodes were positioned at the locations where multiunit activity elicited by flashing visual stimuli was maximal M
At the conclusion of the recording experiments, animals were perfused with fixative (see below) and the brain was sec tioned at 60 jtm on a freezing microtome. The sections were stained with neutral red to distinguish cyto-architectonic bound aries and cortical laminae, which were identified according to the criteria of Henry et al. (1979). Electrolytic lesions (5 j*A by 5 s) made at 0.5- or 1.0-mm intervals during recording ses sions were used to reconstruct electrode tracks, and so identify the laminar position of recorded cells. Figure 1 shows an exam ple of a track with three electrolytic lesions, each separated by 0.5 mm, and also shows the laminar boundaries at the area 17/18 border. The track has run within a single lamina for much of its course (see inset) and the three lesions are contained within lamina 4. Analysis The visual responses of units with a demonstrable transcal losal link were classified into the S, C, B, NO, and associated H categories according to criteria detailed elsewhere (Henry, 1977; Henry et al., 1978). In brief, S cells, like simple cells, have discrete areas of ON or OFF discharge and may receive afferents from either the X or the Y streams. In contrast, both the B and the C cells, like complex cells with uniform receptive fields, fire with composite ON/OFF responses throughout their receptive fields. The two are distinguished from each other in that the B cell has a smaller receptive field, a preference for slower stimuli and has more sustained responses to flashed stim uli. C cells are thought to receive their afferents from the Y stream and the B cells, from the X stream (Henry et al., 1983) The NO cells also respond with composite ON/OFF discharges but fail to show the orientation specificity characteristic of other striate neurons Cells with the H DroDertv like hvDercomulex cells, show a preference for foreshortened stimuli while possessine the orooerties of S C or B cells Units were classified ac
imately 100 for a stimulating pulse of 100-200 fis. Response latency was defined as the shortest latency observed in a clus ter of spikes evoked from 10-20 repetitions of the stimulus. Fig ure 3 shows examples of antidromic and transsynaptic responses to electrical stimulation. In Fig. 3 (upper left), the electrical stimulation is initiated by a naturally occurring spike, which passes down the axon to collide with the electrically initiated spike. In Fig. 3 (lower left), the electrical stimulation is delayed until the natural spike has passed the point of stimulation and the induced antidromic response appears in the record. The ten replicates of the trace in Fig. 3B show the consistency of the re sponse and the absence of jitter in antidromic firing A level of jitter characteristic of trans synaptic responses is apparent in Fig. 3 (uDDer right) Figure 3 (lower right) shows an unusual ex ample where the cell fires with an initial antidromic resDonse followed bv a trans svnaDtic response The utter is small in the cu uy|-pcnonse a. Lidiis-syiidpin. sc.trans c j svnantic uci is MU
Properties of area 17/18 border neurons contributing to the visual transcallosal pathway in the cat
M.E. M c C O U R T , J. T H A L L U R I , AND G . H . H E N R Y Division of Neuroscience, John Curtin School of Medical Research, Australian National University, Canberra, Australia (RECEIVED November 20, 1989; ACCEPTED April 4, 1990)
Abstract In a series of physiological experiments, a total of 203 neurons at the Area 17/18 border were recorded with a callosal link either demonstrated by antidromic or transsynaptic activation from stimulating electrodes located in the homotopic contralateral hemisphere (CH), or in the splenial segment of the corpus callosum (CC). Forty-four percent of the transcallosal cells could also be driven from stimulating electrodes in or just above the lateral geniculate nucleus (OR,). The majority (69%) of transcallosal neurons were classifiable as belonging to the complex family (B and C cells) and most of these were found in the supragranular laminae and in lamina 4A. The ocular dominance distribution of transcallosal cells was trimodal, consisting of roughly equal numbers of monocularly dominated and binocularly balanced neurons. Estimates of conduction time and synaptic delay were obtained for neurons driven from CH, CC, and from OR], and in most instances the response latency was short enough to suggest a monosynaptic input from either the ipsi- or contra-lateral hemisphere. The distribution of transcallosal conduction times showed that S cells, as a class, had significantly faster conduction than cells of the complex family but otherwise there was no obvious signs of multimodality in the distribution curve An analysis of the synaptic delays in transcallosal activation produced a mean of 0.6 to 0.7 ms but some were too short to be consistent with a transsynaptic drive suggesting that some cells with an antidromic drive may have been included in the transsynaptic category Results are interpreted in terms of the contribution made by the corpus callosum to stereoscopic vision. Keywords: Visual cortex, Transcallosal pathways, Receptive fields, Response latency, Cat, Electrical stimulation
Introduction
calls for a revision of the contribution of transcallosal neu rons to coarse stereopsis; whereas they could act alone in the processing of uncrossed-disparity information, they may only supplement the chiasmal contribution in the processing of crossed-disparity signals. The earliest attempts to separate the contribution from the chiasm and the corpus callosum were behavioral experiments which showed that cortical binocularity and stereoscopic depth discrimination in the cat were achieved through both chiasmal and transcallosal pathways (Fox & Blake, 1971; Packwood & Gordon, 1975; Kaye et al., 1981, 1982; Timney et al., 1985; Ptito et al., 1986; Lepore et al., 1986). Several of these studies suggested that the naso-temporal overlap in the chiasm makes a major, although not exclusive, contribution to stereopsis in the cat (Lepore et al., 1986; Ptito et al., 1986). A weak contri bution also comes through the transcallosal pathway so that cats with transsectioned chiasms performed at greater-thanchance levels on disparity-discrimination tasks. The relative weakness of the callosal contribution revealed in these experi ments may provide another indication that crossed-disparity in formation exclusively used in the targets of these studies is Drocessed principally' bv BT or Y retinal Eanslion cells of the naso-temporal overlap
The decussation of the projections arising from each eye results in separate cerebral representations of the two halves of the visual field which are believed to be reunified, in part, through the corpus callosum. In an assessment of this reunion, Bishop and Henry (1971) proposed that the transcallosal linkages were only responsible for coarse stereopsis involving large disparities and that fine stereopsis was mediated through the nasotemporal overlap in the optic chiasm. Subsequent analysis of streaming in the optic chiasm prompted Levick (1977) to sug gest that the partial decussation of the brisk-transient (BT) pop ulation of cat retinal ganglion cells (equivalent to Y-cells in the classification of Enroth-Cugell & Robson, 1966) was well-suited to support crossed-disparity, coarse stereopsis. This proposal
Reprint requests to: G.H. Henry, Centre for Visual Sciences, JCSMR, P.O. Box 334, Canberra, ACT, 2601, Australia. Present address of M.E. McCourt: Department of Psychology and Institute for Neurological Sciences Research, University of Texas (Aus tin), Austin, TX, 78712, USA. Present address of J. Thalluri, School of Applied Science, South Australian College of Advanced Education, Salisbury East, South Aus tralia, 5109, Australia.
83
M.E. McCourt, J. Thalluri, and G.H. Henry
84 In a more direct approach, a variety of experimental proce dures have been employed to assess the physiological response properties of visual neurons contributing to or receiving from the transcallosal pathway. A common technique is to compare neurons in the callosal terminal zone (at the border of areas 17/18) before and after removing the input of the splenial seg ment of the corpus callosum (Payne et al., 1980; Payne et al., 1984; Minciacchi & Antonini, 1984; Cynader et al., 1986; but see Elberger & Smith, 1985, and Payne, 1986 for discussion) or before and after the midline sectioning of the corpus callosum. Both of these intrusions lead to a reduction in the proportion of binocularly activated neurons when compared to the intact cat After splitting the optic chiasm the majority of remaining neurons (viz callosally recipient) which belong to the complex cell familv (B and C cells in the terminoloev of Henrv 1977) possess characteristically laree receDtive fields and are concen irated in the supragranular layers and in upper lamina. 4 (Lepore & Guillemot 1982) A weakness in the ablation and sectioning experiments is the uncertainty that the recorded cells have a link with the corpus callosum. This deficiency is overcome in experiments where the responses of callosal fibers are recorded directly (Hubel & Wiesel, 1967; Berlucchi et al., 1967; Shatz, 1977; Berardi et al., 1987) or where responses to electrical stimulation are used to identify a callosal link (Harvey, 1980; Innocenti, 1980; McCourt et al., 1985, 1987). The direct recordings from fibers of the cor pus callosum reveals that contributing neurons may belong to any response class, that most are binocular and that many possess high-contrast thresholds and low spatial and temporal acuities more typical of cells of the Y-than the X-stream. It is possible however that these experiments using recordings from transcallosal fibers may be contaminated by fibers passing to other visual areas such as the suprasylvian complex (Segraves & Rosenauist 1982a b Rosenauist 1985" Rauschecker et al 1987) where contributing cells are known to have lower visual acuity (DiStephano et aL 1985) m
The application of electrical stimulation to the transcallosal pathway while recording from cells at the border of areas 17/18, the method employed in the present experiments, makes it possible to identify cells projecting to or receiving from the corpus callosum. For this experimental arrangement, neurons contributing to the corpus callosum respond with an antidro mic spike while those receiving a callosal input are driven transsynaptically. Experiments of this type (Harvey, 1980; Innocenti, 1980; McCourt et al., 1985, 1987) have revealed that neurons contributing to and receiving from the corpus callosum belong principally to the complex family (B or C cells), are distributed around the border of laminae 3 and 4, and are predominantly binocular in their responses The present study utilizes the methods of extracellular recording and electrical stimulation to identify callosally projecting and recipient neurons at the border of areas 17/18. The results, therefore, expand and elaborate on the earlier stud ies and more extensive sampling makes it possible to develop more decisive comparisons in the response properties of neurons contributing to and receiving from the corpus callosum. In ad dition, the larger sample of conduction velocities, assessed from response latencies to electrical stimulation at two points in the transcallosal pathway, provides a clearer picture of the time for spike initiation and synaptic delay in the corpus callosum. In a new approach, the present study also compares conduction properties of axons of the transcallosal pathway against those
in afferent fibers coming from the lateral geniculate nucleus (LGN) to the same contributing cell. Another comparison looks at the laminar distributions for cells labeled retrogradely with the axonal tracer, wheat germ agglutinin-horseradish peroxidase ( W G A - H R P ) , and relates it to the distributions of cells con tributing to and receiving from the corpus callosum. The results of the experiments are then interpreted in terms of the contri bution made to the processing of information on coarse stere opsis by signals passing through the transcallosal pathway.
Methods Subjects Twelve adult cats were studied; six were used in the electrophys iological experiments, and six for the tracing of anatomical pathways. The animals ranged from 2.5-3.5 kg in weight. All cats were healthy, possessed normal interocular alignment, and displayed no visual defects.
Electrophysiological
experiments
Preparation Surgery was performed under halothane anesthesia (4% in duction; 1% maintenance dosage) in a mixture of N 0 : 0 (70:30). A long-lasting local anesthetic (2% marcaine) was in jected around the margins of all surgical incisions. The admin istration of halothane was discontinued following surgery, and was replaced during recording sessions by the nitrous oxide to oxygen (70:30) mixture, supplemented by continuous intrave nous infusion of pentobarbitone sodium (1.0 mg k g " h " Nembutal, Bomac Laboratories, Australia; Hammond, 1978). The N O : 0 mixture, delivered through a tracheal cannula, was adjusted around a stroke volume of 40 ml, and a stroke rate of 20 m i n , to achieve a constant end-tidal C 0 level of 4 % . Temperature was measured by a rectal thermoprobe, and was held constant at 38°C with a thermostatically controlled heating blanket. Heart rate, electrocardiogram, and the electro encephalogram were monitored throughout. Paralysis of the ex traocular muscles was achieved by continuous intravenous infusion of pancuronium bromide (1.4 mg k g h Pavulon, N.V. Organon Oss, Holland) and gallamine triethiodide (5 mg kg h Flaxedil, May & Baker, Australia) in depolyte (Intervet Australia Pty Ltd ) A bilateral cervical sympathectomy was performed to further reduce the occurrence of residual eye movements Contact lenses of zero power were installed to pro tect the corneas and the evelids and nictitating membranes were retracted bv 'the tonical aoDlication of phenylephedrine HC1 (10% neosvnDhrine). The nuoils were dilated and accommoda 2
2
1
2
1
2
- 1
2
- 1
- 1
_ 1
_ 1
tinn was rplaxpH hv thp tnniral annliratinn of 1 QTh atronine sul n h » J
t
and T
H o n ih
rnrn d i 7 T e r m
o
,
?
a
T
a
\
r t , f L ^ n.mik h
T
i
of :
w e r e fitted
7e7« f„„H
and
. r a m
centered on tne optic axis wun tne a m ot a zxiss tunaus cam era, wntcn was also used to DacK-project retinal « l°P" tic disc, area centralis) onto a tangent screen at irequem intervals, ine animal was retinoscopicatly retracted and correct ing lenses were fitted accordingly. Small craniotomies were performed for the insertion of the recording and stimulating electrodes. Single units were isolated and their activity was recorded with extracellular glass-coated tungsten microelectrodes (Merrill & Ainsworth, 1972), advanced i a n a m a r K
85
Properties of area 17/18 border neurons in steps of 1 m by a hydraulic microdrive. Stimulating elec trodes were also of tungsten-in-glass construction, but has ex posed tips of approximately 1 mm. Stimulating electrodes were positioned as follows: one in the visual cortex of the contralat eral hemisphere (CH), near the area 17/18 border, at HorsleyClarke coordinates. AP:-3.0, ML: 1.5; a second in the splenium of the corpus callosum (CC) at Horsley-Clarke coordinates AP:3.0 and 5.0, ML:0; and one located in or just above the ip silateral LGN (OR!) at Horsley-Clarke coordinates AP.6.8, ML:9.0. The final position for the stimulating electrode at CH was determined by first probing with a multiunit recording elec trode in CH until units were encountered in retinotopic corre spondence with those isolated by the recording electrode in the iDsilateral cortex The CC and OR stimulating electrodes were positioned at the locations where multiunit activity elicited by flashing visual stimuli was maximal M
At the conclusion of the recording experiments, animals were perfused with fixative (see below) and the brain was sec tioned at 60 jtm on a freezing microtome. The sections were stained with neutral red to distinguish cyto-architectonic bound aries and cortical laminae, which were identified according to the criteria of Henry et al. (1979). Electrolytic lesions (5 j*A by 5 s) made at 0.5- or 1.0-mm intervals during recording ses sions were used to reconstruct electrode tracks, and so identify the laminar position of recorded cells. Figure 1 shows an exam ple of a track with three electrolytic lesions, each separated by 0.5 mm, and also shows the laminar boundaries at the area 17/18 border. The track has run within a single lamina for much of its course (see inset) and the three lesions are contained within lamina 4. Analysis The visual responses of units with a demonstrable transcal losal link were classified into the S, C, B, NO, and associated H categories according to criteria detailed elsewhere (Henry, 1977; Henry et al., 1978). In brief, S cells, like simple cells, have discrete areas of ON or OFF discharge and may receive afferents from either the X or the Y streams. In contrast, both the B and the C cells, like complex cells with uniform receptive fields, fire with composite ON/OFF responses throughout their receptive fields. The two are distinguished from each other in that the B cell has a smaller receptive field, a preference for slower stimuli and has more sustained responses to flashed stim uli. C cells are thought to receive their afferents from the Y stream and the B cells, from the X stream (Henry et al., 1983) The NO cells also respond with composite ON/OFF discharges but fail to show the orientation specificity characteristic of other striate neurons Cells with the H DroDertv like hvDercomulex cells, show a preference for foreshortened stimuli while possessine the orooerties of S C or B cells Units were classified ac
imately 100 for a stimulating pulse of 100-200 fis. Response latency was defined as the shortest latency observed in a clus ter of spikes evoked from 10-20 repetitions of the stimulus. Fig ure 3 shows examples of antidromic and transsynaptic responses to electrical stimulation. In Fig. 3 (upper left), the electrical stimulation is initiated by a naturally occurring spike, which passes down the axon to collide with the electrically initiated spike. In Fig. 3 (lower left), the electrical stimulation is delayed until the natural spike has passed the point of stimulation and the induced antidromic response appears in the record. The ten replicates of the trace in Fig. 3B show the consistency of the re sponse and the absence of jitter in antidromic firing A level of jitter characteristic of trans synaptic responses is apparent in Fig. 3 (uDDer right) Figure 3 (lower right) shows an unusual ex ample where the cell fires with an initial antidromic resDonse followed bv a trans svnaDtic response The utter is small in the cu uy|-pcnonse a. Lidiis-syiidpin. sc.trans c j svnantic uci is MU
